xen test results on Unisys ES7000 using changeset 13801

xen test results on Unisys ES7000 using changeset 13801

[Krysan, Susan]

[-- Attachment #1.1: Type: text/plain, Size: 1290 bytes --]

Host:  Unisys ES7000/one, x86_64, 16 physical processors, 64 GB RAM

 

xen-unstable changeset 13801 booted with dom0_mem=512M

 

Booted, logged into, and ran SLES top command or Windows task manager in
the following domUs and domVTs (#s 2 through 7 ran simultaneously):

 

1.	32-processor 64-bit SLES10 DomU with 62GB memory 
2.	4-processor 64-bit SLES10 DomU with 16GB memory 
3.	4-procesor 32-bit SLES10 DomVT with 2GB memory 
4.	4-processor 32-bit PAE SLES10 DomVT with 16GB memory 
5.	4-processor 64-bit SLES10 DomVT with 16GB memory 
6.	1-processor Windows XP DomVT with 4GB memory 
7.	1-processor Windows 2003 Server DomVT with 4GB memory 

 

Ran xm-test (non-HVM) on dom0 with the following results:

 

Xm-test timing summary:

  Run Started : Mon, 05 Feb 2007 17:50:45 -0500

  Run Stoped  : Mon, 05 Feb 2007 18:36:21 -0500

Xm-test execution summary:

  PASS:  110

  FAIL:  2

  XPASS: 2

  XFAIL: 1

 

Details:

 

 FAIL: 04_memset_smallmem_pos

         Bug #145 Ballooning DomU too low caused run-away

 

XFAIL: 02_network_local_ping_pos

         ping loopback failed for size 65507. ping eth0 failed for size
65507. 

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

 

 


[-- Attachment #1.2: Type: text/html, Size: 10410 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

xen test results on Unisys ES7000 using changeset 13898

[Krysan, Susan]

[-- Attachment #1.1: Type: text/plain, Size: 1599 bytes --]

Host:  Unisys ES7000/one, x86_64, 16 physical processors, 64 GB RAM

 

xen-unstable changeset 13898 booted with dom0_mem=512M

 

Issue:  When attempt to start a domain for which there is not enough
memory, get "Error: an integer is required" instead of a message
indicating there is not enough memory. See xend.log output attached.

 

Booted, logged into, and ran SLES top command or Windows task manager in
the following domUs and domVTs (#s 2 through 7 ran simultaneously):

 

1.	32-processor 64-bit SLES10 DomU with 62GB memory 
2.	4-processor 64-bit SLES10 DomU with 16GB memory 
3.	4-procesor 32-bit SLES10 DomVT with 2GB memory 
4.	4-processor 32-bit PAE SLES10 DomVT with 16GB memory 
5.	4-processor 64-bit SLES10 DomVT with 16GB memory 
6.	1-processor Windows XP DomVT with 4GB memory 
7.	1-processor Windows 2003 Server DomVT with 4GB memory 

 

Ran xm-test (non-HVM) on dom0 with the following results:

 

Xm-test timing summary:

  Run Started : Mon, 12 Feb 2007 09:50:01 -0500

  Run Stoped  : Mon, 12 Feb 2007 10:36:11 -0500

Xm-test execution summary:

  PASS:  109

  FAIL:  3

  XPASS: 2

  XFAIL: 1

 

 

Details:

 

 FAIL: 04_memset_smallmem_pos

         Bug #145 Ballooning DomU too low caused run-away

 

XFAIL: 02_network_local_ping_pos

         ping loopback failed for size 65507. ping eth0 failed for size
65507.

 

 FAIL: 01_vcpu-pin_basic_pos

         failed to switch VCPU 0 to CPU 1

 

 

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

 

 


[-- Attachment #1.2: Type: text/html, Size: 10909 bytes --]

[-- Attachment #2: xendlog.txt --]
[-- Type: text/plain, Size: 3696 bytes --]

[2007-02-12 08:22:11 4551] ERROR (XendDomainInfo:405) VM start failed
Traceback (most recent call last):
  File "/home/unisys/xen-unstable.hg/dist/install/usr/lib64/python/xen/xend/XendDomainInfo.py", line 392, in start
    XendTask.log_progress(31, 60, self._initDomain)
  File "/home/unisys/xen-unstable.hg/dist/install/usr/lib64/python/xen/xend/XendTask.py", line 213, in log_progress
    retval = func(*args, **kwds)
  File "/home/unisys/xen-unstable.hg/dist/install/usr/lib64/python/xen/xend/XendDomainInfo.py", line 1452, in _initDomain
    balloon.free(memory + shadow)
  File "/home/unisys/xen-unstable.hg/dist/install/usr/lib64/python/xen/xend/balloon.py", line 166, in free
    raise VmError(
VmError: I need 16920576 KiB, but dom0_min_mem is 200704 and shrinking to 200704 KiB would leave only 13869888 KiB free.
[2007-02-12 08:22:11 4551] DEBUG (XendDomainInfo:1601) XendDomainInfo.destroy: domid=3
[2007-02-12 08:22:11 4551] DEBUG (XendDomainInfo:1612) XendDomainInfo.destroyDomain(3)
[2007-02-12 08:22:11 4551] DEBUG (image:565) hvm shutdown watch unregistered
[2007-02-12 08:22:11 4551] DEBUG (image:604) hvm reboot feature watch unregistered
[2007-02-12 08:22:11 4551] ERROR (XendDomainInfo:80) Domain construction failed
Traceback (most recent call last):
  File "/home/unisys/xen-unstable.hg/dist/install/usr/lib64/python/xen/xend/XendDomainInfo.py", line 78, in create
    vm.start()
  File "/home/unisys/xen-unstable.hg/dist/install/usr/lib64/python/xen/xend/XendDomainInfo.py", line 392, in start
    XendTask.log_progress(31, 60, self._initDomain)
  File "/home/unisys/xen-unstable.hg/dist/install/usr/lib64/python/xen/xend/XendTask.py", line 213, in log_progress
    retval = func(*args, **kwds)
  File "/home/unisys/xen-unstable.hg/dist/install/usr/lib64/python/xen/xend/XendDomainInfo.py", line 1452, in _initDomain
    balloon.free(memory + shadow)
  File "/home/unisys/xen-unstable.hg/dist/install/usr/lib64/python/xen/xend/balloon.py", line 166, in free
    raise VmError(
VmError: I need 16920576 KiB, but dom0_min_mem is 200704 and shrinking to 200704 KiB would leave only 13869888 KiB free.
[2007-02-12 08:22:11 4551] DEBUG (XendDomainInfo:1601) XendDomainInfo.destroy: domid=None
[2007-02-12 08:22:11 4551] ERROR (xmlrpclib2:232) Internal error handling xend.domain.create
Traceback (most recent call last):
  File "/home/unisys/xen-unstable.hg/dist/install/usr/lib64/python/xen/util/xmlrpclib2.py", line 201, in _marshaled_dispatch
    response = self._dispatch(method, params)
  File "/usr/lib64/python2.4/SimpleXMLRPCServer.py", line 406, in _dispatch
    return func(*params)
  File "/home/unisys/xen-unstable.hg/dist/install/usr/lib64/python/xen/xend/server/XMLRPCServer.py", line 69, in domain_create
    info = XendDomain.instance().domain_create(config)
  File "/home/unisys/xen-unstable.hg/dist/install/usr/lib64/python/xen/xend/XendDomain.py", line 876, in domain_create
    dominfo = XendDomainInfo.create(config)
  File "/home/unisys/xen-unstable.hg/dist/install/usr/lib64/python/xen/xend/XendDomainInfo.py", line 81, in create
    vm.destroy()
  File "/home/unisys/xen-unstable.hg/dist/install/usr/lib64/python/xen/xend/XendDomainInfo.py", line 1603, in destroy
    paths = self._prepare_phantom_paths()
  File "/home/unisys/xen-unstable.hg/dist/install/usr/lib64/python/xen/xend/XendDomainInfo.py", line 1573, in _prepare_phantom_paths
    t = xstransact("%s/device/vbd" % GetDomainPath(self.domid))
  File "/home/unisys/xen-unstable.hg/dist/install/usr/lib64/python/xen/xend/xenstore/xsutil.py", line 26, in GetDomainPath
    return xshandle().get_domain_path(domid)
TypeError: an integer is required

[-- Attachment #3: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

xen test results on Unisys ES7000 using changeset 14006

[Krysan, Susan]

[-- Attachment #1.1: Type: text/plain, Size: 1369 bytes --]

Host:  Unisys ES7000/one, x86_64, 16 physical processors, 64 GB RAM

 

xen-unstable changeset 14006 booted with dom0_mem=512M

 

Booted, logged into, and ran SLES top command or Windows task manager in
the following domUs and domVTs (#s 2 through 7 ran simultaneously):

 

1.	32-processor 64-bit SLES10 DomU with 62GB memory 
2.	4-processor 64-bit SLES10 DomU with 16GB memory 
3.	4-procesor 32-bit SLES10 DomVT with 2GB memory 
4.	4-processor 32-bit PAE SLES10 DomVT with 16GB memory 
5.	4-processor 64-bit SLES10 DomVT with 16GB memory 
6.	1-processor Windows XP DomVT with 4GB memory 
7.	1-processor Windows 2003 Server DomVT with 4GB memory 

 

Ran xm-test (non-HVM) on dom0 with the following results:

 

Xm-test timing summary:

  Run Started : Tue, 20 Feb 2007 10:13:23 -0500

  Run Stoped  : Tue, 20 Feb 2007 10:59:23 -0500

Xm-test execution summary:

  PASS:  109

  FAIL:  3

  XPASS: 1

  XFAIL: 2

 

Details:

 

 FAIL: 04_memset_smallmem_pos

         Bug #145 Ballooning DomU too low caused run-away

 

XFAIL: 02_network_local_ping_pos

         ping loopback failed for size 65507. ping eth0 failed for size
65507.

 

XFAIL: 11_network_domU_ping_pos

         Ping failed for size 65507.

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

 

 


[-- Attachment #1.2: Type: text/html, Size: 10212 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

xen test results on Unisys ES7000 using changeset 14142

[Krysan, Susan]

[-- Attachment #1.1: Type: text/plain, Size: 1310 bytes --]

Host:  Unisys ES7000/one, x86_64, 16 physical processors, 64 GB RAM

 

xen-unstable changeset 14142 booted with dom0_mem=512M

 

Issues: None

 

Booted, logged into, and ran SLES top command or Windows task manager in
the following domUs and domVTs (#s 2 through 7 ran simultaneously):

 

1.	32-processor 64-bit SLES10 DomU with 62GB memory 
2.	4-processor 64-bit SLES10 DomU with 16GB memory 
3.	4-procesor 32-bit SLES10 DomVT with 2GB memory 
4.	4-processor 32-bit PAE SLES10 DomVT with 16GB memory 
5.	4-processor 64-bit SLES10 DomVT with 16GB memory 
6.	1-processor Windows XP DomVT with 4GB memory 
7.	1-processor Windows 2003 Server DomVT with 4GB memory 

 

Ran xm-test (non-HVM) on dom0 with the following results:

 

Xm-test timing summary:

  Run Started : Tue, 27 Feb 2007 06:42:45 -0500

  Run Stoped  : Tue, 27 Feb 2007 07:28:53 -0500

Xm-test execution summary:

  PASS:  110

  FAIL:  2

  XPASS: 2

  XFAIL: 1

 

Details:

 

 FAIL: 04_memset_smallmem_pos

         Bug #145 Ballooning DomU too low caused run-away

 

XFAIL: 02_network_local_ping_pos

         ping loopback failed for size 65507. ping eth0 failed for size
65507.

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

 

 


[-- Attachment #1.2: Type: text/html, Size: 9826 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

xen test results on Unisys ES7000 using changeset 14201

[Krysan, Susan]

[-- Attachment #1.1: Type: text/plain, Size: 1315 bytes --]

Host:  Unisys ES7000/one, x86_64, 16 physical processors, 64 GB RAM

 

xen-unstable changeset 14201 booted with dom0_mem=512M

 

Issues: None

 

Booted, logged into, and ran SLES top command or Windows task manager in
the following domUs and domVTs (#s 2 through 7 ran simultaneously):

 

1.	32-processor 64-bit SLES10 DomU with 62GB memory 
2.	4-processor 64-bit SLES10 DomU with 16GB memory 
3.	4-procesor 32-bit SLES10 DomVT with 2GB memory 
4.	4-processor 32-bit PAE SLES10 DomVT with 16GB memory 
5.	4-processor 64-bit SLES10 DomVT with 16GB memory 
6.	1-processor Windows XP DomVT with 4GB memory 
7.	1-processor Windows 2003 Server DomVT with 4GB memory 

 

Ran xm-test (non-HVM) on dom0 with the following results:

 

Xm-test timing summary:

  Run Started : Tue, 06 Mar 2007 02:18:40 -0500

  Run Stoped  : Tue, 06 Mar 2007 03:04:45 -0500

Xm-test execution summary:

  PASS:  110

  FAIL:  2

  XPASS: 2

  XFAIL: 1

 

 

Details:

 

 FAIL: 04_memset_smallmem_pos

         Bug #145 Ballooning DomU too low caused run-away

 

XFAIL: 02_network_local_ping_pos

         ping loopback failed for size 65507. ping eth0 failed for size
65507.

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

 

 


[-- Attachment #1.2: Type: text/html, Size: 10122 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

xen test results on Unisys ES7000 using changeset 14462

[Krysan, Susan]

[-- Attachment #1.1: Type: text/plain, Size: 1480 bytes --]

Host:  Unisys ES7000/one, x86_64, 16 physical processors, 64 GB RAM

 

xen-unstable changeset 14462 booted with dom0_mem=512M

 

ISSUES: boot of 32-bit SLE10 DomVTs hang.  Serial console and qemu log
provide no indication why.

 

Attempted to boot, log into, and run SLES top command or Windows task
manager in the following domUs and domVTs (with #s 2 through 7 running
simultaneously):

 

1.	32-processor 64-bit SLES10 DomU with 62GB memory 
2.	4-processor 64-bit SLES10 DomU with 16GB memory 
3.	4-procesor 32-bit SLES10 DomVT with 2GB memory 
4.	4-processor 32-bit PAE SLES10 DomVT with 16GB memory 
5.	4-processor 64-bit SLES10 DomVT with 16GB memory 
6.	1-processor Windows XP DomVT with 4GB memory 
7.	1-processor Windows 2003 Server DomVT with 4GB memory 

 

Results:

Domains #3 and #4 hang during boot.

Rest of tests successful.

 

Ran xm-test (non-HVM) on dom0 with the following results:

 

Xm-test timing summary:

  Run Started : Wed, 21 Mar 2007 03:50:38 -0400

  Run Stoped  : Wed, 21 Mar 2007 04:36:35 -0400

Xm-test execution summary:

  PASS:  110

  FAIL:  2

  XPASS: 1

  XFAIL: 2

 

Details:

 

XFAIL: 02_network_local_ping_pos

         ping loopback failed for size 65507. ping eth0 failed for size
65507.

 

XFAIL: 11_network_domU_ping_pos

         Ping failed for size 65507.

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

 

 


[-- Attachment #1.2: Type: text/html, Size: 10929 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

xen test results on Unisys ES7000 using changeset 14552

[Krysan, Susan]

[-- Attachment #1.1: Type: text/plain, Size: 2031 bytes --]

Host:  Unisys ES7000/one, x86_64, 16 physical processors, 64 GB RAM

 

xen-unstable changeset 14552 booted with dom0_mem=512M

 

ISSUES: boot of 32-bit SLES10 DomVTs hang.  Serial console, qemu log,
and xend.log provide no indication why; the 32-bit SLES10 DomVTs boot
when I back up to changeset 14201 (which I ran successfully 2 weeks
ago); the problem was introduced somewhere between changesets 14201 and
14462 (which I ran last week).  Suspect changesets 14391 and 14392.

 

Testing includes running xm-test and also attempting to boot and run
programs in the following domUs and domVTs (running domains #s 3 through
8 simultaneously):

 

1.       32-processor 64-bit SLES10 DomU with 62GB memory - ran cloops,
bonnie, and kernbench

2.       32-processor 64-bit SLES10 DomVT with 62GB memory

3.       4-processor 64-bit SLES10 DomU with 16GB memory  - ran
kernbench

4.       4-procesor 32-bit SLES10 DomVT with 2GB memory - doesn't boot

5.       4-processor 32-bit PAE SLES10 DomVT with 16GB memory - doesn't
boot

6.       4-processor 64-bit SLES10 DomVT with 16GB memory - ran
kernbench

7.       1-processor Windows XP DomVT with 4GB memory - ran 100% cpu
intensive program

8.       1-processor Windows 2003 Server DomVT with 4GB memory - ran
100% cpu intensive program

 

Results:

Domains #4 and #5 do not boot. They hang on SUSE Linux Enterprise Server
screen before login prompt appears.

Opened bug # 940.

Rest of tests successful.

 

Ran xm-test (non-HVM) on dom0 with the following results:

 

Xm-test timing summary:

  Run Started : Mon, 26 Mar 2007 08:27:15 -0400

  Run Stoped  : Mon, 26 Mar 2007 09:12:58 -0400

Xm-test execution summary:

  PASS:  111

  FAIL:  1

  XPASS: 2

  XFAIL: 1

 

Details:

 

XFAIL: 02_network_local_ping_pos

         ping loopback failed for size 65507. ping eth0 failed for size
65507.

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

 

 


[-- Attachment #1.2: Type: text/html, Size: 13936 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

xen test results on Unisys ES7000 using changeset 14756

[Krysan, Susan]

[-- Attachment #1.1: Type: text/plain, Size: 2229 bytes --]

Host:  Unisys ES7000/one, x86_64, 16 physical processors, 64 GB RAM

 

xen-unstable changeset 14756 booted with dom0_mem=512M

 

ISSUES: boot of 32-bit SLES10 DomVTs hang.  Serial console, qemu log,
and xend.log provide no indication why; the 32-bit SLES10 DomVTs boot
when I back up to changeset 14201 (which I ran successfully 2 weeks
ago); the problem was introduced somewhere between changesets 14201 and
14462 (which I ran last week).  Suspect changesets 14391 and 14392.

 

Testing includes running xm-test and also attempting to boot and run
programs in the following domUs and domVTs (running domains #s 3 through
8 simultaneously):

 

1.       32-processor 64-bit SLES10 DomU with 62GB memory - run cloops,
bonnie, and kernbench

2.       32-processor 64-bit SLES10 DomVT with 62GB memory - run
kernbench

3.       4-processor 64-bit SLES10 DomU with 16GB memory  - run
kernbench

4.       4-procesor 32-bit SLES10 DomVT with 2GB memory - doesn't boot

5.       4-processor 32-bit PAE SLES10 DomVT with 16GB memory - doesn't
boot

6.       4-processor 64-bit SLES10 DomVT with 16GB memory - run
kernbench

7.       1-processor Windows XP DomVT with 4GB memory - run 100% cpu
intensive program

8.       1-processor Windows 2003 Server DomVT with 4GB memory - run
100% cpu intensive program

 

Results:

Domain #2 boots with only 8 processors instead of 32

Domains #4 and #5 (32-bit) do not boot. Bug # 940 still applies to #4.
#5 gets kernel panic - not syncing: attempted to kill init

Domain #8 - cannot log in because cannot access qemu console ctrl-alt-2
(or 1 or 3) to send ctrl-alt-delete key sequence

Rest of tests successful.

 

Ran xm-test (non-HVM) on dom0 with the following results:

 

Xm-test timing summary:

  Run Started : Fri, 06 Apr 2007 15:12:33 -0400

  Run Stoped  : Fri, 06 Apr 2007 15:58:06 -0400

Xm-test execution summary:

  PASS:  111

  FAIL:  1

  XPASS: 2

  XFAIL: 1

 

Details:

 

XFAIL: 02_network_local_ping_pos

         ping loopback failed for size 65507. ping eth0 failed for size
65507.

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

 

 


[-- Attachment #1.2: Type: text/html, Size: 14009 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

test results on Unisys ES7000 61x 64gb using 3.0.5 rc3 (c/s 14934) - no new issues

[Krysan, Susan]

[-- Attachment #1.1: Type: text/plain, Size: 1803 bytes --]

Host:  Unisys ES7000/one, x86_64, 16 physical processors, 64 GB RAM

 

xen-unstable changeset 14934 booted with dom0_mem=512M

 

ISSUES: still cannot boot of 32-bit SLES10 DomVTs - bug #940; I'm
currently working on trying to narrow down which c/s caused the break

 

Testing includes running xm-test and also attempting to boot and run
programs in the following domUs and domVTs (running domains #s 3 through
8 simultaneously):

 

1.       32-processor 64-bit SLES10 DomU with 62GB memory - run
kernbench

2.       32-processor 64-bit SLES10 DomVT with 62GB memory - run
kernbench

3.       4-processor 64-bit SLES10 DomU with 16GB memory - run kernbench

4.       4-procesor 32-bit SLES10 DomVT with 2GB memory - doesn't boot

5.       4-processor 32-bit PAE SLES10 DomVT with 16GB memory - doesn't
boot

6.       4-processor 64-bit SLES10 DomVT with 16GB memory - run
kernbench

7.       1-processor Windows XP DomVT with 4GB memory - run 100% cpu
intensive program

8.       1-processor Windows 2003 Server DomVT with 4GB memory - run
100% cpu intensive program

 

Results:

Domains #4 and #5 (32-bit) do not boot.  Bug # 940

Rest ran successfully.

 

Ran xm-test (PV) on dom0 with the following results:

 

Xm-test timing summary:

  Run Started : Wed, 25 Apr 2007 14:04:21 -0400

  Run Stoped  : Wed, 25 Apr 2007 14:50:41 -0400

Xm-test execution summary:

  PASS:  110

  FAIL:  2

  XPASS: 1

  XFAIL: 2

 

Details:

 

XFAIL: 02_network_local_ping_pos

         ping loopback failed for size 65507. ping eth0 failed for size
65507.

 

XFAIL: 11_network_domU_ping_pos

         Ping failed for size 65507.

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

 

 


[-- Attachment #1.2: Type: text/html, Size: 14203 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

test results on Unisys ES7000 32x 128gb using xen-3.1-testing (c/s 15006) - no new issues

[Krysan, Susan]

[-- Attachment #1.1: Type: text/plain, Size: 1933 bytes --]

Host:  Unisys ES7000/one, x86_64, 32 physical processors, 128 GB RAM

 

Xen-3.1-testing changeset 15006 booted with dom0_mem=512M

 

ISSUES: still cannot boot of 32-bit SLES10 DomVTs - bug #940; narrowed
down to c/s 14436

 

Testing includes running xm-test and also attempting to boot and run
programs in the following domUs and domVTs (running domains #s 3 through
9 simultaneously):

 

1.       32-processor 64-bit SLES10 DomU with 62GB memory - run
kernbench optimal load

2.       32-processor 64-bit SLES10 DomVT with 62GB memory - run
kernbench optimal load

3.       4-processor 64-bit SLES10 DomU with 16GB memory - run kernbench
optimal load

4.       4-procesor 32-bit SLES10 DomVT with 2GB memory - doesn't boot

5.       4-processor 32-bit PAE SLES10 DomVT with 16GB memory - doesn't
boot

6.       4-processor 64-bit SLES10 DomVT with 16GB memory - run
kernbench optimal load

7.       1-processor Windows XP DomVT with 4GB memory - run 100% cpu
intensive program

8.       1-processor Windows 2003 Server DomVT with 4GB memory - run
100% cpu intensive program

            9.   8-processor Windows 2003 Enterprise Edition DomVT with
8GB memory - run 100% cpu intensive program

 

Results:

Domains #4 and #5 (32-bit) do not boot.  Bug # 940

Rest ran successfully.

 

Ran xm-test (PV) on dom0 with the following results:

 

Xm-test timing summary:

  Run Started : Thu, 03 May 2007 09:23:44 -0400

  Run Stoped  : Thu, 03 May 2007 10:10:50 -0400

Xm-test execution summary:

  PASS:  110

  FAIL:  2

  XPASS: 1

  XFAIL: 2

 

Details:

 

XFAIL: 02_network_local_ping_pos

         ping loopback failed for size 65507. ping eth0 failed for size
65507.

 

XFAIL: 11_network_domU_ping_pos

         Ping failed for size 65507.

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

 

 


[-- Attachment #1.2: Type: text/html, Size: 14701 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

RE: test results on Unisys ES7000 32x 128gb usingxen-3.1-testing (c/s 15006) - no new issues

[Nakajima, Jun]

[-- Attachment #1.1: Type: text/plain, Size: 2405 bytes --]

Can you try to boot the SLES10 with "nohpet" in the grub command line?

Jun
---
Intel Open Source Technology Center 


________________________________

From: xen-devel-bounces@lists.xensource.com
[mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Krysan,
Susan
Sent: Friday, May 04, 2007 7:30 AM
To: xen-devel@lists.xensource.com
Subject: [Xen-devel] test results on Unisys ES7000 32x 128gb
usingxen-3.1-testing (c/s 15006) - no new issues



Host:  Unisys ES7000/one, x86_64, 32 physical processors, 128 GB RAM

 

Xen-3.1-testing changeset 15006 booted with dom0_mem=512M

 

ISSUES: still cannot boot of 32-bit SLES10 DomVTs - bug #940; narrowed
down to c/s 14436

 

Testing includes running xm-test and also attempting to boot and run
programs in the following domUs and domVTs (running domains #s 3 through
9 simultaneously):

 

1.       32-processor 64-bit SLES10 DomU with 62GB memory - run
kernbench optimal load

2.       32-processor 64-bit SLES10 DomVT with 62GB memory - run
kernbench optimal load

3.       4-processor 64-bit SLES10 DomU with 16GB memory - run kernbench
optimal load

4.       4-procesor 32-bit SLES10 DomVT with 2GB memory - doesn't boot

5.       4-processor 32-bit PAE SLES10 DomVT with 16GB memory - doesn't
boot

6.       4-processor 64-bit SLES10 DomVT with 16GB memory - run
kernbench optimal load

7.       1-processor Windows XP DomVT with 4GB memory - run 100% cpu
intensive program

8.       1-processor Windows 2003 Server DomVT with 4GB memory - run
100% cpu intensive program

            9.   8-processor Windows 2003 Enterprise Edition DomVT with
8GB memory - run 100% cpu intensive program

 

Results:

Domains #4 and #5 (32-bit) do not boot.  Bug # 940

Rest ran successfully.

 

Ran xm-test (PV) on dom0 with the following results:

 

Xm-test timing summary:

  Run Started : Thu, 03 May 2007 09:23:44 -0400

  Run Stoped  : Thu, 03 May 2007 10:10:50 -0400

Xm-test execution summary:

  PASS:  110

  FAIL:  2

  XPASS: 1

  XFAIL: 2

 

Details:

 

XFAIL: 02_network_local_ping_pos

         ping loopback failed for size 65507. ping eth0 failed for size
65507.

 

XFAIL: 11_network_domU_ping_pos

         Ping failed for size 65507.

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

 

 


[-- Attachment #1.2: Type: text/html, Size: 14730 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

RE: test results on Unisys ES7000 32x 128gb usingxen-3.1-testing (c/s 15006) - no new issues

[Krysan, Susan]

[-- Attachment #1.1: Type: text/plain, Size: 2961 bytes --]

Thanks for your response.

I edited the kernel line to append "nohpet" for both the SLES10 hvms
(32-bit and 32-bit pae) and they both hang in the blue SUSE linux
screen.

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

________________________________

From: Nakajima, Jun [mailto:jun.nakajima@intel.com] 
Sent: Friday, May 04, 2007 11:02 AM
To: Krysan, Susan; xen-devel@lists.xensource.com
Subject: RE: [Xen-devel] test results on Unisys ES7000 32x 128gb
usingxen-3.1-testing (c/s 15006) - no new issues

 

Can you try to boot the SLES10 with "nohpet" in the grub command line?

Jun
---
Intel Open Source Technology Center 

 

________________________________

From: xen-devel-bounces@lists.xensource.com
[mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Krysan,
Susan
Sent: Friday, May 04, 2007 7:30 AM
To: xen-devel@lists.xensource.com
Subject: [Xen-devel] test results on Unisys ES7000 32x 128gb
usingxen-3.1-testing (c/s 15006) - no new issues

Host:  Unisys ES7000/one, x86_64, 32 physical processors, 128 GB RAM

 

Xen-3.1-testing changeset 15006 booted with dom0_mem=512M

 

ISSUES: still cannot boot of 32-bit SLES10 DomVTs - bug #940; narrowed
down to c/s 14436

 

Testing includes running xm-test and also attempting to boot and run
programs in the following domUs and domVTs (running domains #s 3 through
9 simultaneously):

 

1.       32-processor 64-bit SLES10 DomU with 62GB memory - run
kernbench optimal load

2.       32-processor 64-bit SLES10 DomVT with 62GB memory - run
kernbench optimal load

3.       4-processor 64-bit SLES10 DomU with 16GB memory - run kernbench
optimal load

4.       4-procesor 32-bit SLES10 DomVT with 2GB memory - doesn't boot

5.       4-processor 32-bit PAE SLES10 DomVT with 16GB memory - doesn't
boot

6.       4-processor 64-bit SLES10 DomVT with 16GB memory - run
kernbench optimal load

7.       1-processor Windows XP DomVT with 4GB memory - run 100% cpu
intensive program

8.       1-processor Windows 2003 Server DomVT with 4GB memory - run
100% cpu intensive program

            9.   8-processor Windows 2003 Enterprise Edition DomVT with
8GB memory - run 100% cpu intensive program

 

Results:

Domains #4 and #5 (32-bit) do not boot.  Bug # 940

Rest ran successfully.

 

Ran xm-test (PV) on dom0 with the following results:

 

Xm-test timing summary:

  Run Started : Thu, 03 May 2007 09:23:44 -0400

  Run Stoped  : Thu, 03 May 2007 10:10:50 -0400

Xm-test execution summary:

  PASS:  110

  FAIL:  2

  XPASS: 1

  XFAIL: 2

 

Details:

 

XFAIL: 02_network_local_ping_pos

         ping loopback failed for size 65507. ping eth0 failed for size
65507.

 

XFAIL: 11_network_domU_ping_pos

         Ping failed for size 65507.

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

 

 


[-- Attachment #1.2: Type: text/html, Size: 17763 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

RE: test results on Unisys ES7000 32x 128gbusingxen-3.1-testing (c/s 15006) - no new issues

[Cui, Dexuan]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1.1: Type: text/plain; charset="gb2312", Size: 3467 bytes --]

Then can you try to boot the ia32/ia32pae SLES10 HVM guest with "hpet=disable" in the grub command line? :)

I guess this can work.

 

-- Dexuan

 

________________________________

From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Krysan, Susan
Sent: 2007Äê5ÔÂ5ÈÕ 0:57
To: Nakajima, Jun; xen-devel@lists.xensource.com
Subject: RE: [Xen-devel] test results on Unisys ES7000 32x 128gbusingxen-3.1-testing (c/s 15006) - no new issues

 

Thanks for your response.

I edited the kernel line to append ¡°nohpet¡± for both the SLES10 hvms (32-bit and 32-bit pae) and they both hang in the blue SUSE linux screen.

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

________________________________

From: Nakajima, Jun [mailto:jun.nakajima@intel.com] 
Sent: Friday, May 04, 2007 11:02 AM
To: Krysan, Susan; xen-devel@lists.xensource.com
Subject: RE: [Xen-devel] test results on Unisys ES7000 32x 128gb usingxen-3.1-testing (c/s 15006) - no new issues

 

Can you try to boot the SLES10 with "nohpet" in the grub command line?

Jun
---
Intel Open Source Technology Center 

 

________________________________

From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Krysan, Susan
Sent: Friday, May 04, 2007 7:30 AM
To: xen-devel@lists.xensource.com
Subject: [Xen-devel] test results on Unisys ES7000 32x 128gb usingxen-3.1-testing (c/s 15006) - no new issues

Host:  Unisys ES7000/one, x86_64, 32 physical processors, 128 GB RAM

 

Xen-3.1-testing changeset 15006 booted with dom0_mem=512M

 

ISSUES: still cannot boot of 32-bit SLES10 DomVTs ¨C bug #940; narrowed down to c/s 14436

 

Testing includes running xm-test and also attempting to boot and run programs in the following domUs and domVTs (running domains #s 3 through 9 simultaneously):

 

1.       32-processor 64-bit SLES10 DomU with 62GB memory ¨C run kernbench optimal load

2.       32-processor 64-bit SLES10 DomVT with 62GB memory ¨C run kernbench optimal load

3.       4-processor 64-bit SLES10 DomU with 16GB memory - run kernbench optimal load

4.       4-procesor 32-bit SLES10 DomVT with 2GB memory ¨C doesn¡¯t boot

5.       4-processor 32-bit PAE SLES10 DomVT with 16GB memory ¨C doesn¡¯t boot

6.       4-processor 64-bit SLES10 DomVT with 16GB memory ¨C run kernbench optimal load

7.       1-processor Windows XP DomVT with 4GB memory ¨C run 100% cpu intensive program

8.       1-processor Windows 2003 Server DomVT with 4GB memory ¨C run 100% cpu intensive program

            9.   8-processor Windows 2003 Enterprise Edition DomVT with 8GB memory ¨C run 100% cpu intensive program

 

Results:

Domains #4 and #5 (32-bit) do not boot.  Bug # 940

Rest ran successfully.

 

Ran xm-test (PV) on dom0 with the following results:

 

Xm-test timing summary:

  Run Started : Thu, 03 May 2007 09:23:44 -0400

  Run Stoped  : Thu, 03 May 2007 10:10:50 -0400

Xm-test execution summary:

  PASS:  110

  FAIL:  2

  XPASS: 1

  XFAIL: 2

 

Details:

 

XFAIL: 02_network_local_ping_pos

         ping loopback failed for size 65507. ping eth0 failed for size 65507.

 

XFAIL: 11_network_domU_ping_pos

         Ping failed for size 65507.

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

 

 


[-- Attachment #1.2: Type: text/html, Size: 22345 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

RE: test results on Unisys ES7000 32x 128gbusingxen-3.1-testing (c/s 15006) - no new issues

[Krysan, Susan]

[-- Attachment #1.1: Type: text/plain, Size: 3865 bytes --]

Dexuan

I was able to successfully boot the ia32 and ia32pae SLES10 HVM guests with "hpet=disable" parameter.

Please let me know if you need any more information or if I can test any patches for you.

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

________________________________

From: Cui, Dexuan [mailto:dexuan.cui@intel.com] 
Sent: Monday, May 07, 2007 12:37 PM
To: Krysan, Susan; Nakajima, Jun; xen-devel@lists.xensource.com
Subject: RE: [Xen-devel] test results on Unisys ES7000 32x 128gbusingxen-3.1-testing (c/s 15006) - no new issues

 

Then can you try to boot the in the grub command line? :)

I guess this can work.

 

-- Dexuan

 

________________________________

From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Krysan, Susan
Sent: 2007年5月5日 0:57
To: Nakajima, Jun; xen-devel@lists.xensource.com
Subject: RE: [Xen-devel] test results on Unisys ES7000 32x 128gbusingxen-3.1-testing (c/s 15006) - no new issues

 

Thanks for your response.

I edited the kernel line to append “nohpet” for both the SLES10 hvms (32-bit and 32-bit pae) and they both hang in the blue SUSE linux screen.

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

________________________________

From: Nakajima, Jun [mailto:jun.nakajima@intel.com] 
Sent: Friday, May 04, 2007 11:02 AM
To: Krysan, Susan; xen-devel@lists.xensource.com
Subject: RE: [Xen-devel] test results on Unisys ES7000 32x 128gb usingxen-3.1-testing (c/s 15006) - no new issues

 

Can you try to boot the SLES10 with "nohpet" in the grub command line?

Jun
---
Intel Open Source Technology Center 

 

________________________________

From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Krysan, Susan
Sent: Friday, May 04, 2007 7:30 AM
To: xen-devel@lists.xensource.com
Subject: [Xen-devel] test results on Unisys ES7000 32x 128gb usingxen-3.1-testing (c/s 15006) - no new issues

Host:  Unisys ES7000/one, x86_64, 32 physical processors, 128 GB RAM

 

Xen-3.1-testing changeset 15006 booted with dom0_mem=512M

 

ISSUES: still cannot boot of 32-bit SLES10 DomVTs - bug #940; narrowed down to c/s 14436

 

Testing includes running xm-test and also attempting to boot and run programs in the following domUs and domVTs (running domains #s 3 through 9 simultaneously):

 

1.       32-processor 64-bit SLES10 DomU with 62GB memory - run kernbench optimal load

2.       32-processor 64-bit SLES10 DomVT with 62GB memory - run kernbench optimal load

3.       4-processor 64-bit SLES10 DomU with 16GB memory - run kernbench optimal load

4.       4-procesor 32-bit SLES10 DomVT with 2GB memory - doesn’t boot

5.       4-processor 32-bit PAE SLES10 DomVT with 16GB memory - doesn’t boot

6.       4-processor 64-bit SLES10 DomVT with 16GB memory - run kernbench optimal load

7.       1-processor Windows XP DomVT with 4GB memory - run 100% cpu intensive program

8.       1-processor Windows 2003 Server DomVT with 4GB memory - run 100% cpu intensive program

            9.   8-processor Windows 2003 Enterprise Edition DomVT with 8GB memory - run 100% cpu intensive program

 

Results:

Domains #4 and #5 (32-bit) do not boot.  Bug # 940

Rest ran successfully.

 

Ran xm-test (PV) on dom0 with the following results:

 

Xm-test timing summary:

  Run Started : Thu, 03 May 2007 09:23:44 -0400

  Run Stoped  : Thu, 03 May 2007 10:10:50 -0400

Xm-test execution summary:

  PASS:  110

  FAIL:  2

  XPASS: 1

  XFAIL: 2

 

Details:

 

XFAIL: 02_network_local_ping_pos

         ping loopback failed for size 65507. ping eth0 failed for size 65507.

 

XFAIL: 11_network_domU_ping_pos

         Ping failed for size 65507.

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

 

 


[-- Attachment #1.2: Type: text/html, Size: 23871 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

RE: test results on Unisys ES7000 32x 128gbusingxen-3.1-testing (c/s 15006) - no new issues

[Cui, Dexuan]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1.1: Type: text/plain; charset="gb2312", Size: 4631 bytes --]

Susan,

Can you try the latest Xen-unstable with/without change set 14436: 0611355e0df2 that lowers the frequency of the HPET DM?

Can you try another platform (i.e. Bensley or something) against the issue? 

 

We have no Unisys ES7000, and we can¡¯t reproduce the issue in Bensley. :(

 

-- Dexuan

 

________________________________

From: Krysan, Susan [mailto:KRYSANS@unisys.com] 
Sent: 2007Äê5ÔÂ10ÈÕ 2:28
To: Cui, Dexuan; Nakajima, Jun; xen-devel@lists.xensource.com
Subject: RE: [Xen-devel] test results on Unisys ES7000 32x 128gbusingxen-3.1-testing (c/s 15006) - no new issues

 

Dexuan

I was able to successfully boot the ia32 and ia32pae SLES10 HVM guests with "hpet=disable" parameter.

Please let me know if you need any more information or if I can test any patches for you.

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

________________________________

From: Cui, Dexuan [mailto:dexuan.cui@intel.com] 
Sent: Monday, May 07, 2007 12:37 PM
To: Krysan, Susan; Nakajima, Jun; xen-devel@lists.xensource.com
Subject: RE: [Xen-devel] test results on Unisys ES7000 32x 128gbusingxen-3.1-testing (c/s 15006) - no new issues

 

Then can you try to boot the in the grub command line? :)

I guess this can work.

 

-- Dexuan

 

________________________________

From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Krysan, Susan
Sent: 2007Äê5ÔÂ5ÈÕ 0:57
To: Nakajima, Jun; xen-devel@lists.xensource.com
Subject: RE: [Xen-devel] test results on Unisys ES7000 32x 128gbusingxen-3.1-testing (c/s 15006) - no new issues

 

Thanks for your response.

I edited the kernel line to append ¡°nohpet¡± for both the SLES10 hvms (32-bit and 32-bit pae) and they both hang in the blue SUSE linux screen.

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

________________________________

From: Nakajima, Jun [mailto:jun.nakajima@intel.com] 
Sent: Friday, May 04, 2007 11:02 AM
To: Krysan, Susan; xen-devel@lists.xensource.com
Subject: RE: [Xen-devel] test results on Unisys ES7000 32x 128gb usingxen-3.1-testing (c/s 15006) - no new issues

 

Can you try to boot the SLES10 with "nohpet" in the grub command line?

Jun
---
Intel Open Source Technology Center 

 

________________________________

From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Krysan, Susan
Sent: Friday, May 04, 2007 7:30 AM
To: xen-devel@lists.xensource.com
Subject: [Xen-devel] test results on Unisys ES7000 32x 128gb usingxen-3.1-testing (c/s 15006) - no new issues

Host:  Unisys ES7000/one, x86_64, 32 physical processors, 128 GB RAM

 

Xen-3.1-testing changeset 15006 booted with dom0_mem=512M

 

ISSUES: still cannot boot of 32-bit SLES10 DomVTs ¨C bug #940; narrowed down to c/s 14436

 

Testing includes running xm-test and also attempting to boot and run programs in the following domUs and domVTs (running domains #s 3 through 9 simultaneously):

 

1.       32-processor 64-bit SLES10 DomU with 62GB memory ¨C run kernbench optimal load

2.       32-processor 64-bit SLES10 DomVT with 62GB memory ¨C run kernbench optimal load

3.       4-processor 64-bit SLES10 DomU with 16GB memory - run kernbench optimal load

4.       4-procesor 32-bit SLES10 DomVT with 2GB memory ¨C doesn¡¯t boot

5.       4-processor 32-bit PAE SLES10 DomVT with 16GB memory ¨C doesn¡¯t boot

6.       4-processor 64-bit SLES10 DomVT with 16GB memory ¨C run kernbench optimal load

7.       1-processor Windows XP DomVT with 4GB memory ¨C run 100% cpu intensive program

8.       1-processor Windows 2003 Server DomVT with 4GB memory ¨C run 100% cpu intensive program

            9.   8-processor Windows 2003 Enterprise Edition DomVT with 8GB memory ¨C run 100% cpu intensive program

 

Results:

Domains #4 and #5 (32-bit) do not boot.  Bug # 940

Rest ran successfully.

 

Ran xm-test (PV) on dom0 with the following results:

 

Xm-test timing summary:

  Run Started : Thu, 03 May 2007 09:23:44 -0400

  Run Stoped  : Thu, 03 May 2007 10:10:50 -0400

Xm-test execution summary:

  PASS:  110

  FAIL:  2

  XPASS: 1

  XFAIL: 2

 

Details:

 

XFAIL: 02_network_local_ping_pos

         ping loopback failed for size 65507. ping eth0 failed for size 65507.

 

XFAIL: 11_network_domU_ping_pos

         Ping failed for size 65507.

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

 

 


[-- Attachment #1.2: Type: text/html, Size: 29763 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

RE: test results on Unisys ES7000 32x 128gbusingxen-3.1-testing (c/s 15006) - no new issues

[Krysan, Susan]

[-- Attachment #1.1: Type: text/plain, Size: 5069 bytes --]

When I back out changeset 14436, the ia32 and ia32pae hvms boot fine.  When I add that changeset back in, they do no boot.

The only other system I have available to try is a Dell with VT enabled.   I can try to get access to it .

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

________________________________

From: Cui, Dexuan [mailto:dexuan.cui@intel.com] 
Sent: Wednesday, May 09, 2007 10:38 PM
To: Krysan, Susan; xen-devel@lists.xensource.com
Subject: RE: [Xen-devel] test results on Unisys ES7000 32x 128gbusingxen-3.1-testing (c/s 15006) - no new issues

 

Susan,

Can you try the latest Xen-unstable with/without change set 14436: 0611355e0df2 that lowers the frequency of the HPET DM?

Can you try another platform (i.e. Bensley or something) against the issue? 

 

We have no Unisys ES7000, and we can’t reproduce the issue in Bensley. :(

 

-- Dexuan

 

________________________________

From: Krysan, Susan [mailto:KRYSANS@unisys.com] 
Sent: 2007年5月10日 2:28
To: Cui, Dexuan; Nakajima, Jun; xen-devel@lists.xensource.com
Subject: RE: [Xen-devel] test results on Unisys ES7000 32x 128gbusingxen-3.1-testing (c/s 15006) - no new issues

 

Dexuan

I was able to successfully boot the ia32 and ia32pae SLES10 HVM guests with "hpet=disable" parameter.

Please let me know if you need any more information or if I can test any patches for you.

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

________________________________

From: Cui, Dexuan [mailto:dexuan.cui@intel.com] 
Sent: Monday, May 07, 2007 12:37 PM
To: Krysan, Susan; Nakajima, Jun; xen-devel@lists.xensource.com
Subject: RE: [Xen-devel] test results on Unisys ES7000 32x 128gbusingxen-3.1-testing (c/s 15006) - no new issues

 

Then can you try to boot the in the grub command line? :)

I guess this can work.

 

-- Dexuan

 

________________________________

From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Krysan, Susan
Sent: 2007年5月5日 0:57
To: Nakajima, Jun; xen-devel@lists.xensource.com
Subject: RE: [Xen-devel] test results on Unisys ES7000 32x 128gbusingxen-3.1-testing (c/s 15006) - no new issues

 

Thanks for your response.

I edited the kernel line to append “nohpet” for both the SLES10 hvms (32-bit and 32-bit pae) and they both hang in the blue SUSE linux screen.

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

________________________________

From: Nakajima, Jun [mailto:jun.nakajima@intel.com] 
Sent: Friday, May 04, 2007 11:02 AM
To: Krysan, Susan; xen-devel@lists.xensource.com
Subject: RE: [Xen-devel] test results on Unisys ES7000 32x 128gb usingxen-3.1-testing (c/s 15006) - no new issues

 

Can you try to boot the SLES10 with "nohpet" in the grub command line?

Jun
---
Intel Open Source Technology Center 

 

________________________________

From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Krysan, Susan
Sent: Friday, May 04, 2007 7:30 AM
To: xen-devel@lists.xensource.com
Subject: [Xen-devel] test results on Unisys ES7000 32x 128gb usingxen-3.1-testing (c/s 15006) - no new issues

Host:  Unisys ES7000/one, x86_64, 32 physical processors, 128 GB RAM

 

Xen-3.1-testing changeset 15006 booted with dom0_mem=512M

 

ISSUES: still cannot boot of 32-bit SLES10 DomVTs - bug #940; narrowed down to c/s 14436

 

Testing includes running xm-test and also attempting to boot and run programs in the following domUs and domVTs (running domains #s 3 through 9 simultaneously):

 

1.       32-processor 64-bit SLES10 DomU with 62GB memory - run kernbench optimal load

2.       32-processor 64-bit SLES10 DomVT with 62GB memory - run kernbench optimal load

3.       4-processor 64-bit SLES10 DomU with 16GB memory - run kernbench optimal load

4.       4-procesor 32-bit SLES10 DomVT with 2GB memory - doesn’t boot

5.       4-processor 32-bit PAE SLES10 DomVT with 16GB memory - doesn’t boot

6.       4-processor 64-bit SLES10 DomVT with 16GB memory - run kernbench optimal load

7.       1-processor Windows XP DomVT with 4GB memory - run 100% cpu intensive program

8.       1-processor Windows 2003 Server DomVT with 4GB memory - run 100% cpu intensive program

            9.   8-processor Windows 2003 Enterprise Edition DomVT with 8GB memory - run 100% cpu intensive program

 

Results:

Domains #4 and #5 (32-bit) do not boot.  Bug # 940

Rest ran successfully.

 

Ran xm-test (PV) on dom0 with the following results:

 

Xm-test timing summary:

  Run Started : Thu, 03 May 2007 09:23:44 -0400

  Run Stoped  : Thu, 03 May 2007 10:10:50 -0400

Xm-test execution summary:

  PASS:  110

  FAIL:  2

  XPASS: 1

  XFAIL: 2

 

Details:

 

XFAIL: 02_network_local_ping_pos

         ping loopback failed for size 65507. ping eth0 failed for size 65507.

 

XFAIL: 11_network_domU_ping_pos

         Ping failed for size 65507.

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

 

 


[-- Attachment #1.2: Type: text/html, Size: 30618 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

test results on Unisys ES7000 32x 128gb using xen-3.1-testing (c/s 15040) - no new issues

[Krysan, Susan]

[-- Attachment #1.1: Type: text/plain, Size: 1855 bytes --]

Host:  Unisys ES7000/one, x86_64, 32 physical processors, 128 GB RAM

 

Xen-3.1-testing changeset 15040 booted with dom0_mem=512M

 

ISSUES: still cannot boot of 32-bit SLES10 DomVTs - bug #940; narrowed
down to c/s 14436

 

Testing includes running xm-test and also attempting to boot and run
programs in the following domUs and domVTs (running domains #s 3 through
9 simultaneously):

 

1.       32-processor 64-bit SLES10 DomU with 62GB memory - run
kernbench optimal load

2.       32-processor 64-bit SLES10 DomVT with 62GB memory - run
kernbench optimal load

3.       4-processor 64-bit SLES10 DomU with 16GB memory - run kernbench
optimal load

4.       4-procesor 32-bit SLES10 DomVT with 2GB memory - doesn't boot

5.       4-processor 32-bit PAE SLES10 DomVT with 16GB memory - doesn't
boot

6.       4-processor 64-bit SLES10 DomVT with 16GB memory - run
kernbench optimal load

7.       1-processor Windows XP DomVT with 4GB memory - run 100% cpu
intensive program

8.       1-processor Windows 2003 Server DomVT with 4GB memory - run
100% cpu intensive program

            9.   8-processor Windows 2003 Enterprise Edition DomVT with
8GB memory - run 100% cpu intensive program

 

Results:

Domains #4 and #5 (32-bit) do not boot.  Bug # 940

Rest ran successfully.

 

Ran xm-test (DomU) on dom0 with the following results:

 

Xm-test timing summary:

  Run Started : Mon, 14 May 2007 14:04:02 -0400

  Run Stoped  : Mon, 14 May 2007 14:51:11 -0400

Xm-test execution summary:

  PASS:  111

  FAIL:  1

  XPASS: 2

  XFAIL: 1

 

Details:

 

XFAIL: 02_network_local_ping_pos

         ping loopback failed for size 65507. ping eth0 failed for size
65507.

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

 

 


[-- Attachment #1.2: Type: text/html, Size: 14257 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

RE: can't boot 32-bit SLES10 HVMs since c/s 14436 wasintroduced

[Krysan, Susan]

[-- Attachment #1.1: Type: text/plain, Size: 3822 bytes --]

Hi.  Dexuan.  Just a note to let you know that I also cannot boot 32-bit SLES10 hvms on Dell PowerEdge SC1430.  I can only boot if I add the hpet=disable parameter.  Also wanted to let you know that as I keep testing this issue, the failure to boot actually happens in a few different ways, on both the ES7000 and the Dell: sometimes with the kernel panic: Badness in smp_call_function at arch/i386/kernel/smp.c:595, sometimes the boot hangs in blue screen before login prompt appears, and sometimes it gets to the login prompt but the screen flashes off and on and when enter the userid, I keep getting a message that the login timed out after 60 seconds and I am unable to login.

 

I will be using the Dell box to keep decreasing the TSC_PER_HPET_TICK value as you suggest to see if I can find a value that will work.  Then I will try that value on the Unisys ES7000.  

 

Could you explain what exactly this value controls and why it is needed? (since I can boot successfully without it)

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

________________________________

From: Cui, Dexuan [mailto:dexuan.cui@intel.com] 
Sent: Monday, May 14, 2007 9:57 PM
To: Krysan, Susan
Subject: RE: [Xen-devel] test results on Unisys ES7000 32x 128gb usingxen-3.1-testing (c/s 15040) - no new issues

 

Hi Susan,

Sorry to see the issue “32-bit SLES10 HVM guest can’t boot on Unisys ES7000” is still there. :-(

 

Can you try changing the macro in xen/arch/x86/hvm/hpet.c? i.e., change it to 16, 8, 4, 2, 1, and see if the guest can boot.

I think at least changing the macro to 1 can boot the guest definitely since this means disabling the patch actually.

 

-- Dexuan

 

________________________________

From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Krysan, Susan
Sent: 2007年5月15日 5:15
To: xen-devel@lists.xensource.com
Subject: [Xen-devel] test results on Unisys ES7000 32x 128gb usingxen-3.1-testing (c/s 15040) - no new issues

 

Host:  Unisys ES7000/one, x86_64, 32 physical processors, 128 GB RAM

 

Xen-3.1-testing changeset 15040 booted with dom0_mem=512M

 

ISSUES: still cannot boot of 32-bit SLES10 DomVTs - bug #940; narrowed down to c/s 14436

 

Testing includes running xm-test and also attempting to boot and run programs in the following domUs and domVTs (running domains #s 3 through 9 simultaneously):

 

1.       32-processor 64-bit SLES10 DomU with 62GB memory - run kernbench optimal load

2.       32-processor 64-bit SLES10 DomVT with 62GB memory - run kernbench optimal load

3.       4-processor 64-bit SLES10 DomU with 16GB memory - run kernbench optimal load

4.       4-procesor 32-bit SLES10 DomVT with 2GB memory - doesn’t boot

5.       4-processor 32-bit PAE SLES10 DomVT with 16GB memory - doesn’t boot

6.       4-processor 64-bit SLES10 DomVT with 16GB memory - run kernbench optimal load

7.       1-processor Windows XP DomVT with 4GB memory - run 100% cpu intensive program

8.       1-processor Windows 2003 Server DomVT with 4GB memory - run 100% cpu intensive program

            9.   8-processor Windows 2003 Enterprise Edition DomVT with 8GB memory - run 100% cpu intensive program

 

Results:

Domains #4 and #5 (32-bit) do not boot.  Bug # 940

Rest ran successfully.

 

Ran xm-test (DomU) on dom0 with the following results:

 

Xm-test timing summary:

  Run Started : Mon, 14 May 2007 14:04:02 -0400

  Run Stoped  : Mon, 14 May 2007 14:51:11 -0400

Xm-test execution summary:

  PASS:  111

  FAIL:  1

  XPASS: 2

  XFAIL: 1

 

Details:

 

XFAIL: 02_network_local_ping_pos

         ping loopback failed for size 65507. ping eth0 failed for size 65507.

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

 

 


[-- Attachment #1.2: Type: text/html, Size: 22511 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

Test results on Unisys ES7000 32x 128gb using xen-unstable (c/s 15080) - 1 new issue

[Krysan, Susan]

[-- Attachment #1.1: Type: text/plain, Size: 2099 bytes --]

Host:  Unisys ES7000/one, x86_64, 32 physical processors, 128 GB RAM

 

xen-unstable changeset 15080 booted with dom0_mem=512M

 

ISSUES:

New: host halts upon shutdown of 126000mb domU and domVT

Old: must specify hpet=disable kernel parameter to get 32-bit SLES10
domVTs to boot - bug #940; narrowed down to c/s 14436

 

Testing includes running xm-test and also attempting to boot and run
programs in the following domUs and domVTs (running domains #s 3 through
9 simultaneously):

 

1.       32-processor 64-bit SLES10 domU with 126gb (126000mb) memory -
run kernbench optimal load

2.       32-processor 64-bit SLES10 domVT with 126gb (126000mb) memory -
run kernbench optimal load

3.       4-processor 64-bit SLES10 domU with 16gb memory - run kernbench
optimal load

4.       4-procesor 32-bit SLES10 domVT with 2gb memory - doesn't boot

5.       4-processor 32-bit PAE SLES10 domVT with 16gb memory - doesn't
boot

6.       4-processor 64-bit SLES10 domVT with 16gb memory - run
kernbench optimal load

7.       1-processor Windows XP domVT with 4gb memory - run 100% cpu
intensive program

8.       1-processor Windows 2003 Server domVT with 4gb memory - run
100% cpu intensive program

            9.   8-processor Windows 2003 Enterprise Edition domVT with
8gb memory - run 100% cpu intensive program

 

Results:

Domains #4 and #5 (32-bit) do not boot.  Bug # 940

Domain #1 and #2 - host crashes upon shutdown of domU (reducing the
memory in config file to 124000mb works fine)

Rest ran successfully.

 

Ran xm-test (DomU) on dom0 with the following results:

 

Xm-test timing summary:

  Run Started : Wed, 30 May 2007 08:02:22 -0400

  Run Stoped  : Wed, 30 May 2007 08:49:15 -0400

Xm-test execution summary:

  PASS:  111

  FAIL:  1

  XPASS: 2

  XFAIL: 1

 

Details:

 

XFAIL: 02_network_local_ping_pos

         ping loopback failed for size 65507. ping eth0 failed for size
65507.

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

 

 


[-- Attachment #1.2: Type: text/html, Size: 15017 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

Test results on Unisys ES7000 32x 128gb using xen-unstable (c/s 15470) - 2 old issues; 2 new issue

[Krysan, Susan]

[-- Attachment #1.1: Type: text/plain, Size: 3544 bytes --]

Host:  Unisys ES7000/one, x86_64, 32 physical processors, 128 GB RAM

 

xen-unstable changeset 15470 booted with dom0_mem=512M

 

2 NEW ISSUES:

Networking does not work when a default gateway is not specified

All xm-test tests fail with XmTestLib.NetConfig.NetworkError: Failed to
show vif0.0 aliases: 65280 - workaround is to comment out the call to
cleanDom0Aliases in xm-test/lib/XmTestLib/NetConfig.py

 

2 OLD ISSUES:

Host halts upon shutdown of 126000mb domU and domVT

Must specify hpet=disable kernel parameter to get 32-bit SLES10 domVTs
to boot - bug #940; narrowed down to c/s 14436

 

Testing includes running xm-test and also attempting to boot and run
programs in the following domUs and domVTs (running domains #s 3 through
9 simultaneously):

 

1.       32-processor 64-bit SLES10 domU with 126gb (126000mb) memory -
run kernbench optimal load

2.       32-processor 64-bit SLES10 domVT with 126gb (126000mb) memory -
run kernbench optimal load

3.       4-processor 64-bit SLES10 domU with 16gb memory - run kernbench
optimal load

4.       4-procesor 32-bit SLES10 domVT with 2gb memory - doesn't boot

5.       4-processor 32-bit PAE SLES10 domVT with 16gb memory - doesn't
boot

6.       4-processor 64-bit SLES10 domVT with 16gb memory - run
kernbench optimal load

7.       1-processor Windows XP domVT with 4gb memory - run 100% cpu
intensive program

8.       1-processor Windows 2003 Server domVT with 4gb memory - run
100% cpu intensive program

            9.   8-processor Windows 2003 Enterprise Edition domVT with
8gb memory - run 100% cpu intensive program

 

Results:

 

Domains #4 and #5 (32-bit) do not boot without hpet=disable parameter.
Bug # 940

Domain #1 and #2 - able to run kernbench in these domains, but host
crashes when shut them down; serport does not provide a reason why.  

Reducing the memory of these domains to 124000mb used to work, but have
tested 122000mb and 120000mb and host still halts. Still running tests
to determine the largest size domain that works.

 

The other 5 tests ran successfully.

 

Ran xm-test (DomU) on dom0 with the following results:

 

Could only run after commenting out the call to cleanDom0Aliases in
xm-test/lib/XmTestLib/NetConfig.py

 

Xm-test timing summary:

  Run Started : Sat, 07 Jul 2007 12:28:04 -0400

  Run Stoped  : Sat, 07 Jul 2007 12:59:19 -0400

Xm-test execution summary:

  PASS:  100

  FAIL:  12

  XPASS: 0

  XFAIL: 3

 

Details:

 

 FAIL: 05_block_attach_and_dettach_device_repeatedly_pos

         xm block-attach returned invalid 256 != 0

 

 FAIL: 09_block_attach_and_dettach_device_check_data_pos

         xm block-attach returned invalid 256 != 0

 

 FAIL: 13_create_multinic_pos

        Unknown reason

 

XFAIL: 02_network_local_ping_pos

        Unknown reason

 

 FAIL: 03_network_local_tcp_pos

        Unknown reason

 

 FAIL: 04_network_local_udp_pos

        Unknown reason

 

XFAIL: 05_network_dom0_ping_pos

        Unknown reason

 

 FAIL: 06_network_dom0_tcp_pos

        Unknown reason

 

 FAIL: 07_network_dom0_udp_pos

        Unknown reason

 

XFAIL: 11_network_domU_ping_pos

        Unknown reason

 

 FAIL: 12_network_domU_tcp_pos

        Unknown reason

 

 FAIL: 13_network_domU_udp_pos

        Unknown reason

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

 

 


[-- Attachment #1.2: Type: text/html, Size: 22165 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

RE: Test results on Unisys ES7000 32x 128gb usingxen-unstable (c/s 15470) - 2 old issues; 2 new issue

[Ian Pratt]

> Host:  Unisys ES7000/one, x86_64, 32 physical processors, 128 GB RAM
> 
> 2 OLD ISSUES:
> 
> Host halts upon shutdown of 126000mb domU and domVT
> 
> Testing includes running xm-test and also attempting to boot and run
> programs in the following domUs and domVTs (running domains #s 3
> through 9 simultaneously):
>  
> 1.       32-processor 64-bit SLES10 domU with 126gb (126000mb) memory
-
> run kernbench optimal load
> 
> 2.       32-processor 64-bit SLES10 domVT with 126gb (126000mb) memory
> - run kernbench optimal load
> 
> 
> Domain #1 and #2 - able to run kernbench in these domains, but host
> crashes when shut them down; serport does not provide a reason why.
> 
> Reducing the memory of these domains to 124000mb used to work, but
have
> tested 122000mb and 120000mb and host still halts. Still running tests
> to determine the largest size domain that works.

Are you saying that booting a single very large 64b guest either VT or
PV and then shutting it down will kill the host?

If you boot a debug build of xen, do you get any messages out the serial
port?

Thanks,
Ian

RE: Test results on Unisys ES7000 32x 128gb usingxen-unstable (c/s 15470) - 2 old issues; 2 new issue

[Krysan, Susan]

Yes, the problem was happening on large memory guests, PV and VT, and I
only tried 64-bit guests with such a large size.

However, the server I was testing on recently had software installed in
dom0 which monitors the servers up time (heartbeat) and when the
heartbeat is lost, the software halts the server.  Large memory guests
take a long time to start and shut down, but during shut down, the host
becomes unresponsive.  With the heartbeat monitoring enabled, this
unresponsiveness was resulting in the software halting the server.  I
disabled this feature and the host no longer halts.

Do you have any ideas as to where we can look to determine why the
server seems to be starved for resources, and thus unable to respond to
the heartbeat, when a large virtual machine is shutting down?  Keep in
mind that we run with dom0_mem=512M because otherwise starting the large
VMs takes much longer.  

Thanks,
Sue Krysan
Linux Systems Group
Unisys Corporation
 
-----Original Message-----
From: Ian Pratt [mailto:Ian.Pratt@cl.cam.ac.uk] 
Sent: Sunday, July 08, 2007 4:36 PM
To: Krysan, Susan; xen-devel@lists.xensource.com
Cc: ian.pratt@cl.cam.ac.uk
Subject: RE: [Xen-devel] Test results on Unisys ES7000 32x 128gb
usingxen-unstable (c/s 15470) - 2 old issues; 2 new issue

> Host:  Unisys ES7000/one, x86_64, 32 physical processors, 128 GB RAM
> 
> 2 OLD ISSUES:
> 
> Host halts upon shutdown of 126000mb domU and domVT
> 
> Testing includes running xm-test and also attempting to boot and run
> programs in the following domUs and domVTs (running domains #s 3
> through 9 simultaneously):
>  
> 1.       32-processor 64-bit SLES10 domU with 126gb (126000mb) memory
-
> run kernbench optimal load
> 
> 2.       32-processor 64-bit SLES10 domVT with 126gb (126000mb) memory
> - run kernbench optimal load
> 
> 
> Domain #1 and #2 - able to run kernbench in these domains, but host
> crashes when shut them down; serport does not provide a reason why.
> 
> Reducing the memory of these domains to 124000mb used to work, but
have
> tested 122000mb and 120000mb and host still halts. Still running tests
> to determine the largest size domain that works.

Are you saying that booting a single very large 64b guest either VT or
PV and then shutting it down will kill the host?

If you boot a debug build of xen, do you get any messages out the serial
port?

Thanks,
Ian

RE: Test results on Unisys ES7000 32x 128gb usingxen-unstable (c/s 15470) - 2 old issues; 2 new issue

[Krysan, Susan]

Also, the serial port and logs give no indication of the problem.  I
will rebuild with debug and try again.

Thanks,
Sue Krysan
Linux Systems Group
Unisys Corporation
 

-----Original Message-----
From: Krysan, Susan 
Sent: Monday, July 09, 2007 10:55 AM
To: 'Ian Pratt'; xen-devel@lists.xensource.com
Subject: RE: [Xen-devel] Test results on Unisys ES7000 32x 128gb
usingxen-unstable (c/s 15470) - 2 old issues; 2 new issue

Yes, the problem was happening on large memory guests, PV and VT, and I
only tried 64-bit guests with such a large size.

However, the server I was testing on recently had software installed in
dom0 which monitors the servers up time (heartbeat) and when the
heartbeat is lost, the software halts the server.  Large memory guests
take a long time to start and shut down, but during shut down, the host
becomes unresponsive.  With the heartbeat monitoring enabled, this
unresponsiveness was resulting in the software halting the server.  I
disabled this feature and the host no longer halts.

Do you have any ideas as to where we can look to determine why the
server seems to be starved for resources, and thus unable to respond to
the heartbeat, when a large virtual machine is shutting down?  Keep in
mind that we run with dom0_mem=512M because otherwise starting the large
VMs takes much longer.  

Thanks,
Sue Krysan
Linux Systems Group
Unisys Corporation
 
-----Original Message-----
From: Ian Pratt [mailto:Ian.Pratt@cl.cam.ac.uk] 
Sent: Sunday, July 08, 2007 4:36 PM
To: Krysan, Susan; xen-devel@lists.xensource.com
Cc: ian.pratt@cl.cam.ac.uk
Subject: RE: [Xen-devel] Test results on Unisys ES7000 32x 128gb
usingxen-unstable (c/s 15470) - 2 old issues; 2 new issue

> Host:  Unisys ES7000/one, x86_64, 32 physical processors, 128 GB RAM
> 
> 2 OLD ISSUES:
> 
> Host halts upon shutdown of 126000mb domU and domVT
> 
> Testing includes running xm-test and also attempting to boot and run
> programs in the following domUs and domVTs (running domains #s 3
> through 9 simultaneously):
>  
> 1.       32-processor 64-bit SLES10 domU with 126gb (126000mb) memory
-
> run kernbench optimal load
> 
> 2.       32-processor 64-bit SLES10 domVT with 126gb (126000mb) memory
> - run kernbench optimal load
> 
> 
> Domain #1 and #2 - able to run kernbench in these domains, but host
> crashes when shut them down; serport does not provide a reason why.
> 
> Reducing the memory of these domains to 124000mb used to work, but
have
> tested 122000mb and 120000mb and host still halts. Still running tests
> to determine the largest size domain that works.

Are you saying that booting a single very large 64b guest either VT or
PV and then shutting it down will kill the host?

If you boot a debug build of xen, do you get any messages out the serial
port?

Thanks,
Ian

Re: Test results on Unisys ES7000 32x 128gb usingxen-unstable (c/s 15470) - 2 old issues; 2 new issue

[Keir Fraser]

This is pretty simple. The domain's memory map is torn down synchronously
when it is killed by domain0, via XEN_DOMCTL_destroydomain. This can take a
long time, and during that time that domain0 vcpu is not interruptible.

The possible fixes are:
 1. Find out which bit of domain_kill() takes the longest time and optimise
it so it takes much less time. Unfortunately it is still going to be
proportional to the memory size of the domain.
 2. Make domain destruction asynchronous (probably by introducing hypervisor
threads, analogous to kernel threads).

Possibly we need both (1) and (2). I don't think we can avoid doing (2) in
the long term, really.

 -- Keir

On 9/7/07 16:01, "Krysan, Susan" <KRYSANS@unisys.com> wrote:

> Also, the serial port and logs give no indication of the problem.  I
> will rebuild with debug and try again.
> 
> Thanks,
> Sue Krysan
> Linux Systems Group
> Unisys Corporation
>  
> 
> -----Original Message-----
> From: Krysan, Susan
> Sent: Monday, July 09, 2007 10:55 AM
> To: 'Ian Pratt'; xen-devel@lists.xensource.com
> Subject: RE: [Xen-devel] Test results on Unisys ES7000 32x 128gb
> usingxen-unstable (c/s 15470) - 2 old issues; 2 new issue
> 
> Yes, the problem was happening on large memory guests, PV and VT, and I
> only tried 64-bit guests with such a large size.
> 
> However, the server I was testing on recently had software installed in
> dom0 which monitors the servers up time (heartbeat) and when the
> heartbeat is lost, the software halts the server.  Large memory guests
> take a long time to start and shut down, but during shut down, the host
> becomes unresponsive.  With the heartbeat monitoring enabled, this
> unresponsiveness was resulting in the software halting the server.  I
> disabled this feature and the host no longer halts.
> 
> Do you have any ideas as to where we can look to determine why the
> server seems to be starved for resources, and thus unable to respond to
> the heartbeat, when a large virtual machine is shutting down?  Keep in
> mind that we run with dom0_mem=512M because otherwise starting the large
> VMs takes much longer.
> 
> Thanks,
> Sue Krysan
> Linux Systems Group
> Unisys Corporation
>  
> -----Original Message-----
> From: Ian Pratt [mailto:Ian.Pratt@cl.cam.ac.uk]
> Sent: Sunday, July 08, 2007 4:36 PM
> To: Krysan, Susan; xen-devel@lists.xensource.com
> Cc: ian.pratt@cl.cam.ac.uk
> Subject: RE: [Xen-devel] Test results on Unisys ES7000 32x 128gb
> usingxen-unstable (c/s 15470) - 2 old issues; 2 new issue
> 
>> Host:  Unisys ES7000/one, x86_64, 32 physical processors, 128 GB RAM
>> 
>> 2 OLD ISSUES:
>> 
>> Host halts upon shutdown of 126000mb domU and domVT
>> 
>> Testing includes running xm-test and also attempting to boot and run
>> programs in the following domUs and domVTs (running domains #s 3
>> through 9 simultaneously):
>>  
>> 1.       32-processor 64-bit SLES10 domU with 126gb (126000mb) memory
> -
>> run kernbench optimal load
>> 
>> 2.       32-processor 64-bit SLES10 domVT with 126gb (126000mb) memory
>> - run kernbench optimal load
>> 
>> 
>> Domain #1 and #2 - able to run kernbench in these domains, but host
>> crashes when shut them down; serport does not provide a reason why.
>> 
>> Reducing the memory of these domains to 124000mb used to work, but
> have
>> tested 122000mb and 120000mb and host still halts. Still running tests
>> to determine the largest size domain that works.
> 
> Are you saying that booting a single very large 64b guest either VT or
> PV and then shutting it down will kill the host?
> 
> If you boot a debug build of xen, do you get any messages out the serial
> port?
> 
> Thanks,
> Ian
> 
> 
> 
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xensource.com
> http://lists.xensource.com/xen-devel

Test results on Unisys ES7000 32x 128gb using xen-unstable (c/s 15521) - 4 old issues

[Krysan, Susan]

[-- Attachment #1.1: Type: text/plain, Size: 3261 bytes --]

Host:  Unisys ES7000/one, x86_64, 32 physical processors, 128 GB RAM

 

xen-unstable changeset 15521 booted with dom0_mem=512M acpi=on numa=on

 

OLD ISSUES:

*	Networking does not work when a default gateway is not specified
*	All xm-test tests fail with XmTestLib.NetConfig.NetworkError:
Failed to show vif0.0 aliases: 65280 - workaround is to comment out the
call to cleanDom0Aliases in xm-test/lib/XmTestLib/NetConfig.py
*	Must specify hpet=disable kernel parameter to get 32-bit SLES10
domVTs to boot - bug #940; narrowed down to c/s 14436
*	Shutdown of large domains takes a long time, during which time
dom0 is not interruptible (due to the synchronous tearing down of the
memory mape

 

Testing includes running xm-test and also attempting to boot and run
programs in the following domUs and domVTs (running domains #s 3 through
9 simultaneously):

 

1.       32-processor 64-bit SLES10 domU with 124gb memory - run
kernbench optimal load

2.       32-processor 64-bit SLES10 domVT with 124gb memory - run
kernbench optimal load

3.       4-processor 64-bit SLES10 domU with 16gb memory - run kernbench
optimal load

4.       4-procesor 32-bit SLES10 domVT with 2gb memory (booted with
hpet=disable) - run kernbench optimal load

5.       4-processor 32-bit PAE SLES10 domVT with 16gb memory (booted
with hpet=disable) - run kernbench optimal load

6.       4-processor 64-bit SLES10 domVT with 16gb memory - run
kernbench optimal load

7.       1-processor Windows XP domVT with 4gb memory - run 100% cpu
intensive program

8.       1-processor Windows 2003 Server domVT with 4gb memory - run
100% cpu intensive program

            9.   8-processor Windows 2003 Enterprise Edition domVT with
16gb memory - run 100% cpu intensive program

 

Results:

 

All domains ran successfully.

 

Ran xm-test (DomU) on dom0 with the following results:

 

(Could only run after commenting out the call to cleanDom0Aliases in
xm-test/lib/XmTestLib/NetConfig.py)

 

Xm-test timing summary:

  Run Started : Mon, 16 Jul 2007 16:07:13 -0400

  Run Stoped  : Mon, 16 Jul 2007 16:38:21 -0400

Xm-test execution summary:

  PASS:  100

  FAIL:  12

  XPASS: 0

  XFAIL: 3

 

Details:

 

 FAIL: 05_block_attach_and_dettach_device_repeatedly_pos

         xm block-attach returned invalid 256 != 0

 

 FAIL: 09_block_attach_and_dettach_device_check_data_pos

         xm block-attach returned invalid 256 != 0

 

 FAIL: 13_create_multinic_pos

        Unknown reason

 

XFAIL: 02_network_local_ping_pos

        Unknown reason

 

 FAIL: 03_network_local_tcp_pos

        Unknown reason

 

 FAIL: 04_network_local_udp_pos

        Unknown reason

 

XFAIL: 05_network_dom0_ping_pos

        Unknown reason

 

 FAIL: 06_network_dom0_tcp_pos

        Unknown reason

 

 FAIL: 07_network_dom0_udp_pos

        Unknown reason

 

XFAIL: 11_network_domU_ping_pos

        Unknown reason

 

FAIL: 12_network_domU_tcp_pos

        Unknown reason

 

 FAIL: 13_network_domU_udp_pos

        Unknown reason

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

 

 


[-- Attachment #1.2: Type: text/html, Size: 22998 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

Test results on Unisys ES7000 32x 128gb using xen-unstable (c/s 15672 and es7000bigsmp patch) - 4 old issues, 0 new issues

[Krysan, Susan]

[-- Attachment #1.1: Type: text/plain, Size: 3587 bytes --]

Host:  Unisys ES7000/one, x86_64, 32 physical processors, 128 GB RAM

 

xen-unstable changeset 15672 booted with dom0_mem=512M acpi=on numa=on
and the es7000bigsmp patch applied (which cleans up ES7000 specific
bootup code - submitted by Raj Subrahmanian)

 

OLD ISSUES:

*	after c/s 15203, networking does not work unless a default
gateway is specified
*	after c/s 15203, all xm-test tests fail with
XmTestLib.NetConfig.NetworkError: Failed to show vif0.0 aliases: 65280 -
workaround is to comment out lines in cleanDom0Aliases in
xm-test/lib/XmTestLib/NetConfig.py
*	bug #940 - Must specify hpet=disable kernel parameter to get
32-bit SLES10 domVTs to boot; narrowed down to c/s 14436 but the patch
originator cannot recreate on his hardware
*	bug #1037 - Shutdown of large domains takes a long time, during
which time dom0 is not interruptible (due to the synchronous tearing
down of the memory map)

 

Testing includes running xm-test and also attempting to boot and run
programs in the following domUs and domVTs (running domains #s 3 through
9 simultaneously):

 

1.       32-processor 64-bit SLES10 domU with 124gb memory - run
kernbench optimal load

2.       32-processor 64-bit SLES10 domVT with 124gb memory - run
kernbench optimal load

3.       4-processor 64-bit SLES10 domU with 16gb memory - run kernbench
optimal load

4.       4-procesor 32-bit SLES10 domVT with 2gb memory (booted with
hpet=disable) - run kernbench optimal load

5.       4-processor 32-bit PAE SLES10 domVT with 16gb memory (booted
with hpet=disable) - run kernbench optimal load

6.       4-processor 64-bit SLES10 domVT with 16gb memory - run
kernbench optimal load

7.       1-processor Windows XP domVT with 4gb memory - run 100% cpu
intensive program

8.       1-processor Windows 2003 Server domVT with 4gb memory - run
100% cpu intensive program

            9.   8-processor Windows 2003 Enterprise Edition domVT with
16gb memory - run 100% cpu intensive program

 

Results:

 

All domains ran successfully.

 

Ran xm-test (DomU) on dom0 with the following results:

 

(Could only run after commenting out the call to cleanDom0Aliases in
xm-test/lib/XmTestLib/NetConfig.py and most of the failures are related
to not finding device vif0.0 - this was not necessary before c/s 15203)

 

Xm-test timing summary:

  Run Started : Tue, 07 Aug 2007 15:57:51 -0400

  Run Stoped  : Tue, 07 Aug 2007 16:29:31 -0400

Xm-test execution summary:

  PASS:  100

  FAIL:  12

  XPASS: 0

  XFAIL: 3

 

Details:

 

 FAIL: 05_block_attach_and_dettach_device_repeatedly_pos

         xm block-attach returned invalid 256 != 0

 

 FAIL: 09_block_attach_and_dettach_device_check_data_pos

         xm block-attach returned invalid 256 != 0

 

 FAIL: 13_create_multinic_pos

        Unknown reason

 

XFAIL: 02_network_local_ping_pos

        Unknown reason

 

 FAIL: 03_network_local_tcp_pos

        Unknown reason

 

 FAIL: 04_network_local_udp_pos

        Unknown reason

 

XFAIL: 05_network_dom0_ping_pos

        Unknown reason

 

 FAIL: 06_network_dom0_tcp_pos

        Unknown reason

 

 FAIL: 07_network_dom0_udp_pos

        Unknown reason

 

XFAIL: 11_network_domU_ping_pos

        Unknown reason

 

 FAIL: 12_network_domU_tcp_pos

        Unknown reason

 

 FAIL: 13_network_domU_udp_pos

        Unknown reason

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

 

 


[-- Attachment #1.2: Type: text/html, Size: 23303 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

Test results on Unisys ES7000 48x 160gb using xen-unstable c/s 15730 - 4 old issues, 1 new issue

[Krysan, Susan]

[-- Attachment #1.1: Type: text/plain, Size: 3475 bytes --]

Host:  Unisys ES7000/one, x86_64, 48 processors, 160 GB RAM

 

xen-unstable changeset 15730 compiled with max_phys_cpus=64 and booted
with dom0_mem=512M numa=on

 

NEW ISSUE:

*	Bug #1049 - When boot the host with more than 166 GB RAM, the
host only sees a maximum of 166GB 

 

OLD ISSUES:

*	Bug #1050 - after c/s 15203, networking does not work unless a
default gateway is specified
*	Bug #1051 - after c/s 15203, all xm-test tests fail with
XmTestLib.NetConfig.NetworkError: Failed to show vif0.0 aliases: 65280 -
workaround is to comment out lines in cleanDom0Aliases in
xm-test/lib/XmTestLib/NetConfig.py until xm-test code is changed to
accommodate the new bridge naming scheme
*	Bug #940 - Must specify hpet=disable kernel parameter to get
32-bit SLES10 domVTs to boot; narrowed down to c/s 14436 but the patch
originator cannot recreate on his hardware
*	Bug #1037 - Shutdown of large domains takes a long time, during
which time dom0 is not interruptible (due to the synchronous tearing
down of the memory map)

 

Testing includes running xm-test and also attempting to boot and run
programs in the following domUs and domVTs (running domains #s 3 through
9 simultaneously):

 

1.       32-processor 64-bit SLES10 domU with 156gb memory - run
kernbench optimal load

2.       32-processor 64-bit SLES10 domVT with 150gb memory - run
kernbench optimal load

3.       4-processor 64-bit SLES10 domU with 16gb memory - run kernbench
optimal load 

4.       4-procesor 32-bit SLES10 domVT with 2gb memory (booted with
hpet=disable) - run kernbench optimal load

5.       4-processor 32-bit PAE SLES10 domVT with 16gb memory (booted
with hpet=disable) - run kernbench optimal load

6.       4-processor 64-bit SLES10 domVT with 16gb memory - run
kernbench optimal load

7.       1-processor Windows XP domVT with 4gb memory - run 100% cpu
intensive program

8.       1-processor Windows 2003 Server domVT with 4gb memory - run
100% cpu intensive program

            9.   8-processor Windows 2003 Enterprise Edition domVT with
16gb memory - run 100% cpu intensive program

 

Results:

 

All domains ran successfully.

 

Ran xm-test (DomU) on dom0 with the following results:

 

(Could only run after commenting out the call to cleanDom0Aliases in
xm-test/lib/XmTestLib/NetConfig.py and most of the failures are related
to not finding device vif0.0 - this was not necessary before c/s 15203)

 

Xm-test timing summary:

  Run Started : Thu, 23 Aug 2007 16:13:35 -0400

  Run Stoped  : Thu, 23 Aug 2007 16:48:29 -0400

Xm-test execution summary:

  PASS:  102

  FAIL:  10

  XPASS: 0

  XFAIL: 3

 

Details:

 

 FAIL: 13_create_multinic_pos

        Unknown reason

 

XFAIL: 02_network_local_ping_pos

        Unknown reason

 

 FAIL: 03_network_local_tcp_pos

        Unknown reason

 

 FAIL: 04_network_local_udp_pos

        Unknown reason

 

XFAIL: 05_network_dom0_ping_pos

        Unknown reason

 

 FAIL: 06_network_dom0_tcp_pos

        Unknown reason

 

 FAIL: 07_network_dom0_udp_pos

        Unknown reason

 

XFAIL: 11_network_domU_ping_pos

        Unknown reason

 

 FAIL: 12_network_domU_tcp_pos

        Unknown reason

 

 FAIL: 13_network_domU_udp_pos

        Unknown reason

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

 

 


[-- Attachment #1.2: Type: text/html, Size: 23310 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

Test results on Unisys ES7000 64x 166gb using xen-unstable c/s 15880 -no new issues, 5 old issues

[Krysan, Susan]

[-- Attachment #1.1: Type: text/plain, Size: 3525 bytes --]

Host:  Unisys ES7000/one, x86_64, 64 processors, 256GB RAM, but xen only
boots with 166 GB RAM due to bug#1049

 

xen-unstable changeset 15880 compiled with max_phys_cpus=64 and booted
with dom0_mem=512M numa=on

 

OLD ISSUES:

 

*	Bug #1050 - after c/s 15203, networking does not work unless a
default gateway is specified
*	Bug #1051 - after c/s 15203, all xm-test tests fail with
XmTestLib.NetConfig.NetworkError: Failed to show vif0.0 aliases: 65280 -
workaround is to comment out lines in cleanDom0Aliases in
xm-test/lib/XmTestLib/NetConfig.py until xm-test code is changed to
accommodate the new bridge naming scheme
*	Bug #1049 - When boot the host with more than 166 GB RAM, the
host only sees a maximum of 166GB 
*	Bug #1037 - Shutdown of large domains takes a long time, during
which time dom0 is not interruptible (due to the synchronous tearing
down of the memory map)
*	Bug #940 - Must specify hpet=disable kernel parameter to get
32-bit SLES10 domVTs to boot; narrowed down to c/s 14436 but the patch
originator cannot recreate on his hardware

 

Testing includes running xm-test and also attempting to boot and run
programs in the following domUs and domVTs (running domains #s 3 through
9 simultaneously):

 

1.       32-processor 64-bit SLES10 domU with 160gb memory - run
kernbench

2.       32-processor 64-bit SLES10 domVT with 160gb memory - run
kernbench

3.       8-processor 64-bit SLES10 domU with 16gb memory - run kernbench


4.       8-procesor 32-bit SLES10 domVT with 4gb memory (booted with
hpet=disable) - run kernbench

5.       8-processor 32-bit PAE SLES10 domVT with 16gb memory (booted
with hpet=disable) - run kernbench

6.       8-processor 64-bit SLES10 domVT with 16gb memory - run
kernbench

7.       1-processor Windows XP domVT with 4gb memory - run 100% cpu
intensive program

8.       1-processor Windows 2003 Server domVT with 4gb memory - run
100% cpu intensive program

            9.   8-processor Windows 2003 Enterprise Edition domVT with
16gb memory - run 100% cpu intensive program

 

Results:

 

All domains ran successfully.

 

Ran xm-test (DomU) on dom0 with the following results:

 

(Could only run after commenting out the call to cleanDom0Aliases in
xm-test/lib/XmTestLib/NetConfig.py and most of the failures are related
to not finding device vif0.0 - this was not necessary before c/s 15203)

 

Xm-test timing summary:

  Run Started : Thu, 13 Sep 2007 10:21:53 -0400

  Run Stoped  : Thu, 13 Sep 2007 10:56:39 -0400

Xm-test execution summary:

  PASS:  101

  FAIL:  11

  XPASS: 0

  XFAIL: 3

 

Details:

 

 FAIL: 13_create_multinic_pos

        Unknown reason

 

XFAIL: 02_network_local_ping_pos

        Unknown reason

 

 FAIL: 03_network_local_tcp_pos

        Unknown reason

 

 FAIL: 04_network_local_udp_pos

        Unknown reason

 

XFAIL: 05_network_dom0_ping_pos

        Unknown reason

 

 FAIL: 06_network_dom0_tcp_pos

        Unknown reason

 

 FAIL: 07_network_dom0_udp_pos

        Unknown reason

 

XFAIL: 11_network_domU_ping_pos

        Unknown reason

 

 FAIL: 12_network_domU_tcp_pos

        Unknown reason

 

 FAIL: 13_network_domU_udp_pos

        Unknown reason

 

 FAIL: 04_restore_withdevices_pos

         Timed out waiting for console command

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

 

 


[-- Attachment #1.2: Type: text/html, Size: 24752 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

Test results on Unisys ES7000 64x 192gb using xen-3.1.1-rc2 c/s 15444 -no new issues, 2 old issues

[Krysan, Susan]

[-- Attachment #1.1: Type: text/plain, Size: 2246 bytes --]

Host:  Unisys ES7000/one, x86_64, 48 processors, 192GB RAM

 

xen-3.1-testing -rc2 changeset 15444 compiled with max_phys_cpus=64 and
booted with dom0_mem=1024M numa=on

 

OLD ISSUES:

 

*	Bug #1037 - Shutdown of large domains takes a long time, during
which time dom0 is not interruptible (due to the synchronous tearing
down of the memory map) - for example, shutdown of 180gb domu causes 5
min 15 sec dom0 unresponsiveness
*	Bug #940 - Must specify hpet=disable kernel parameter to get
32-bit SLES10 domVTs to boot; narrowed down to c/s 14436 but the patch
originator cannot recreate on his hardware

 

Testing includes running xm-test and also attempting to boot and run
programs in the following domUs and domVTs (running domains #s 3 through
1- simultaneously):

 

1.       32-processor 64-bit SLES10 domU with 160gb memory - run
kernbench

2.       32-processor 64-bit SLES10 domVT with 160gb memory - run
kernbench

3.       8-processor 64-bit SLES10 domU with 16gb memory - run kernbench

4.       8-processor 32-bit SLES10 domU with 4gb memory - run kernbench

5.       8-procesor 32-bit SLES10 domVT with 4gb memory (booted with
hpet=disable) - run kernbench

6.       8-processor 32-bit PAE SLES10 domVT with 16gb memory (booted
with hpet=disable) - run kernbench

7.       8-processor 64-bit SLES10 domVT with 16gb memory - run
kernbench

8.       8-processor Windows 2003 Enterprise Edition domVT with 16gb
memory - run 100% cpu intensive program

9.       1-processor Windows XP domVT with 4gb memory - run 100% cpu
intensive program

10.   1-processor Windows 2003 Server domVT with 4gb memory - run 100%
cpu intensive program

 

Results:

 

All tests were successful, except for the 2 old issues above.

 

Xm-test timing summary:

  Run Started : Fri, 28 Sep 2007 18:10:12 -0400

  Run Stoped  : Fri, 28 Sep 2007 19:00:13 -0400

Xm-test execution summary:

  PASS:  111

  FAIL:  1

  XPASS: 2

  XFAIL: 1

 

Details:

 

XFAIL: 02_network_local_ping_pos

         ping loopback failed for size 65507. ping eth0 failed for size
65507.

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

 

 


[-- Attachment #1.2: Type: text/html, Size: 17701 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

Re: Test results on Unisys ES7000 64x 192gb using xen-3.1.1-rc2 c/s 15444 -no new issues, 2 old issues

[Keir Fraser]

[-- Attachment #1.1: Type: text/plain, Size: 509 bytes --]


Do you still see bug #1037 on xen-unstable? There is a fix for this issue
there, but there¹s not much point in backporting it if it doesn¹t yet work.

 -- Keir

On 30/9/07 16:35, "Krysan, Susan" <KRYSANS@unisys.com> wrote:

> OLD ISSUES:
>  
> * Bug #1037 - Shutdown of large domains takes a long time, during which time
> dom0 is not interruptible (due to the synchronous tearing down of the memory
> map) ­ for example, shutdown of 180gb domu causes 5 min 15 sec dom0
> unresponsiveness 



[-- Attachment #1.2: Type: text/html, Size: 1254 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

Test results on Unisys ES7000 48x 192gb using xen-3.1.1-rc3 c/s 15461 -no new issues, 2 old issues

[Krysan, Susan]

[-- Attachment #1.1: Type: text/plain, Size: 2482 bytes --]

Host:  Unisys ES7000/one, x86_64, 48 processors, 192 gb memory

 

xen-3.1-testing -rc3 changeset 15461 compiled with max_phys_cpus=64 and
booted with dom0_mem=1024M and numa=on

 

OLD ISSUES:

 

*	Bug #1037 - Shutdown of large domains takes a long time, during
which time dom0 is not interruptible (due to the synchronous tearing
down of the memory map) - for example, shutdown of 180gb domu causes 5
min 15 sec dom0 unresponsiveness  FYI  the patch submitted with
preemption code does get called but it is not working for us.  Our team
is still looking into it
*	Bug #940 - Must specify hpet=disable kernel parameter to get
32-bit SLES10 domVTs to boot; narrowed down to c/s 14436 but the patch
originator cannot recreate on his hardware

 

TESTS include running xm-test and also attempting to boot and run
programs in the following domUs and domVTs (running domains #s 3 through
10 simultaneously):

 

1.       32-processor 64-bit SLES10 domU with 160gb memory - run
kernbench

2.       32-processor 64-bit SLES10 domVT with 160gb memory - run
kernbench

3.       8-processor 64-bit SLES10 domU with 16gb memory - run kernbench

4.       8-processor 32-bit SLES10 domU with 4gb memory - run kernbench

5.       8-procesor 32-bit SLES10 domVT with 4gb memory (booted with
hpet=disable) - run kernbench

6.       8-processor 32-bit PAE SLES10 domVT with 16gb memory (booted
with hpet=disable) - run kernbench

7.       8-processor 64-bit SLES10 domVT with 16gb memory - run
kernbench

8.       8-processor 64-bit Windows 2003 Enterprise Edition domVT with
16gb memory - run 100% cpu intensive program

9.       2-processor 32-bit Windows XP domVT with 4gb memory - run 100%
cpu intensive program

10.   1-processor 32-bit Windows 2003 Server domVT with 4gb memory - run
100% cpu intensive program

 

RESULTS:

 

All tests were successful, except for the 2 old issues above.

 

Xm-test timing summary:

  Run Started : Tue, 09 Oct 2007 08:03:46 -0400

  Run Stoped  : Tue, 09 Oct 2007 08:54:21 -0400

Xm-test execution summary:

  PASS:  110

  FAIL:  2

  XPASS: 1

  XFAIL: 2

 

Details:

 

XFAIL: 02_network_local_ping_pos

         ping loopback failed for size 65507. ping eth0 failed for size
65507.

 

XFAIL: 11_network_domU_ping_pos

         Ping failed for size 65507.

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

 

 


[-- Attachment #1.2: Type: text/html, Size: 18609 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

Test results on Unisys ES7000 32x 128gb using xen-3.1.2-rc1 c/s 15489 -no new issues, 2 old issues

[Krysan, Susan]

[-- Attachment #1.1: Type: text/plain, Size: 2477 bytes --]

Host:  Unisys ES7000/one, x86_64, 32 processors, 128 gb memory

 

xen-3.1.2 -rc1 changeset 15489 compiled with max_phys_cpus=64 and booted
with dom0_mem=1024M and numa=on

 

OLD ISSUES:

 

*	Bug #1037 - Shutdown of large domains takes a long time, during
which time dom0 is not interruptible (due to the synchronous tearing
down of the memory map) - for example, shutdown of 180gb domu causes 5
min 15 sec dom0 unresponsiveness  FYI  the patch submitted with
preemption code does get called but it is not working for us.  Our team
is still looking into it
*	Bug #940 - Must specify hpet=disable kernel parameter to get
32-bit SLES10 domVTs to boot; narrowed down to c/s 14436 but the patch
originator cannot recreate on his hardware

 

TESTS include running xm-test and also attempting to boot and run
programs in the following domUs and domVTs (running domains #s 3 through
10 simultaneously):

 

1.       32-processor 64-bit SLES10 domU with 124gb memory - run
kernbench

2.       32-processor 64-bit SLES10 domVT with 124gb memory - run
kernbench

3.       8-processor 64-bit SLES10 domU with 16gb memory - run kernbench

4.       8-processor 64-bit SLES10 domVT with 16gb memory - run
kernbench

5.       8-processor 32-bit SLES10 domU with 4gb memory - run kernbench

6.       8-processor 32-bit SLES10 domVT with 4gb memory (booted with
hpet=disable) - run kernbench

7.       8-processor 32-bit PAE SLES10 domVT with 16gb memory (booted
with hpet=disable) - run kernbench

8.       8-processor 64-bit Windows 2003 Enterprise Edition domVT with
16gb memory - run 100% cpu intensive program

9.       2-processor 32-bit Windows XP domVT with 4gb memory - run 100%
cpu intensive program

10.   1-processor 32-bit Windows 2003 Server domVT with 4gb memory - run
100% cpu intensive program

 

RESULTS:

 

All tests were successful, except for the 2 old issues above.

 

Xm-test timing summary:

  Run Started : Wed, 24 Oct 2007 10:43:30 -0400

  Run Stoped  : Wed, 24 Oct 2007 11:35:03 -0400

Xm-test execution summary:

  PASS:  110

  FAIL:  2

  XPASS: 1

  XFAIL: 2

 

Details:

 

XFAIL: 02_network_local_ping_pos

         ping loopback failed for size 65507. ping eth0 failed for size
65507.

 

XFAIL: 11_network_domU_ping_pos

         Ping failed for size 65507.

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

 

 


[-- Attachment #1.2: Type: text/html, Size: 18743 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

Test results on Unisys ES7000 64x 256gb using xen-3.1.2-rc1 c/s 15490 -1 new issue, 2 old issues

[Krysan, Susan]

[-- Attachment #1.1: Type: text/plain, Size: 2840 bytes --]

Host:  Unisys ES7000/one, x86_64, 64 processors, 256 gb memory

 

xen-3.1.2-rc1 changeset 15490 compiled with max_phys_cpus=64 and booted
with dom0_mem=2048M and numa=on

 

OLD ISSUES:

 

*	Bug #1037 - Shutdown of large domains takes a long time, during
which time dom0 is not interruptible (due to the synchronous tearing
down of the memory map) - for example, shutdown of 180gb domu causes 5
min 15 sec dom0 unresponsiveness  
*	Bug #940 - Must specify hpet=disable kernel parameter to get
32-bit SLES10 domVTs to boot; narrowed down to c/s 14436 but the patch
originator cannot recreate on his hardware

 

NEW ISSUE:

*	Xm-test gets this failure, but we had never run it with 256GB
memory before so it may not be a regression:

11_create_concurrent_pos failed with:

Error: (12, 'Cannot allocate memory')

So increased dom0_mem parameter to 4096M, rebooted, and reran but get
the same problem; will test today with less host memory to see if still
fails

 

TESTS include running xm-test and also attempting to boot and run
programs in the following domUs and domVTs (running domains #s 3 through
10 simultaneously):

 

1.       32-processor 64-bit SLES10 domU with 248gb memory - run
kernbench

2.       32-processor 64-bit SLES10 domVT with 248gb memory - run
kernbench

3.       32-processor 64-bit SLES10 domU with 64gb memory - run
kernbench

4.       32-processor 64-bit SLES10 domVT with 64gb memory - run
kernbench

5.       8-processor 32-bit SLES10 domU with 4gb memory - run kernbench

6.       8-processor 32-bit SLES10 domVT with 4gb memory (booted with
hpet=disable) - run kernbench

7.       8-processor 32-bit PAE SLES10 domVT with 16gb memory (booted
with hpet=disable) - run kernbench

8.       8-processor 64-bit Windows 2003 Enterprise Edition domVT with
16gb memory - booted but could not log in because OS needs to be
activated

9.       2-processor 32-bit Windows XP domVT with 4gb memory - run 100%
cpu intensive program

10.   1-processor 32-bit Windows 2003 Server domVT with 4gb memory - run
100% cpu intensive program

 

RESULTS:

 

All tests were successful, except for the 2 old issues above.

 

Xm-test timing summary:

  Run Started : Wed, 31 Oct 2007 17:03:45 -0400

  Run Stoped  : Wed, 31 Oct 2007 17:53:49 -0400

Xm-test execution summary:

  PASS:  109

  FAIL:  3

  XPASS: 1

  XFAIL: 2

 

Details:

 

 FAIL: 11_create_concurrent_pos

         [34] Failed to create domain

 

XFAIL: 02_network_local_ping_pos

         ping loopback failed for size 65507. ping eth0 failed for size
65507.

 

XFAIL: 11_network_domU_ping_pos

         Ping failed for size 65507.

 

 

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

 

 


[-- Attachment #1.2: Type: text/html, Size: 21197 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

UPDATE: Test results on Unisys ES7000 64x 256gb using xen-3.1.2-rc1 c/s 15490 -1 new issue, 2 old issues

[Krysan, Susan]

[-- Attachment #1.1: Type: text/plain, Size: 3568 bytes --]

The RESULTS section in the original message I sent needs to indicate
that there is 1 new issue.

 

Additional info for this new issue, the serial port shows this entry:

  Cannot handle page request order 0!

When the host is booted with half of the resources, 32 processors and
128 gb memory, test 11_create_concurrent_pos of xm-test does not fail.

 

Any ideas what could be causing this to fail when there is ample host
memory?

 

________________________________

From: Krysan, Susan 
Sent: Thursday, November 01, 2007 6:29 AM
To: 'xen-devel@lists.xensource.com'
Subject: Test results on Unisys ES7000 64x 256gb using xen-3.1.2-rc1 c/s
15490 -1 new issue, 2 old issues

 

Host:  Unisys ES7000/one, x86_64, 64 processors, 256 gb memory

 

xen-3.1.2-rc1 changeset 15490 compiled with max_phys_cpus=64 and booted
with dom0_mem=2048M and numa=on

 

OLD ISSUES:

 

*	Bug #1037 - Shutdown of large domains takes a long time, during
which time dom0 is not interruptible (due to the synchronous tearing
down of the memory map) - for example, shutdown of 180gb domu causes 5
min 15 sec dom0 unresponsiveness  
*	Bug #940 - Must specify hpet=disable kernel parameter to get
32-bit SLES10 domVTs to boot; narrowed down to c/s 14436 but the patch
originator cannot recreate on his hardware

 

NEW ISSUE:

*	Xm-test gets this failure, but we had never run it with 256GB
memory before so it may not be a regression:

11_create_concurrent_pos failed with:

Error: (12, 'Cannot allocate memory')

So increased dom0_mem parameter to 4096M, rebooted, and reran but get
the same problem; will test today with less host memory to see if still
fails

 

TESTS include running xm-test and also attempting to boot and run
programs in the following domUs and domVTs (running domains #s 3 through
10 simultaneously):

 

1.       32-processor 64-bit SLES10 domU with 248gb memory - run
kernbench

2.       32-processor 64-bit SLES10 domVT with 248gb memory - run
kernbench

3.       32-processor 64-bit SLES10 domU with 64gb memory - run
kernbench

4.       32-processor 64-bit SLES10 domVT with 64gb memory - run
kernbench

5.       8-processor 32-bit SLES10 domU with 4gb memory - run kernbench

6.       8-processor 32-bit SLES10 domVT with 4gb memory (booted with
hpet=disable) - run kernbench

7.       8-processor 32-bit PAE SLES10 domVT with 16gb memory (booted
with hpet=disable) - run kernbench

8.       8-processor 64-bit Windows 2003 Enterprise Edition domVT with
16gb memory - booted but could not log in because OS needs to be
activated

9.       2-processor 32-bit Windows XP domVT with 4gb memory - run 100%
cpu intensive program

10.   1-processor 32-bit Windows 2003 Server domVT with 4gb memory - run
100% cpu intensive program

 

RESULTS:

 

All tests were successful, except for 1 new issue and 2 old issues (see
above).

 

Xm-test timing summary:

  Run Started : Wed, 31 Oct 2007 17:03:45 -0400

  Run Stoped  : Wed, 31 Oct 2007 17:53:49 -0400

Xm-test execution summary:

  PASS:  109

  FAIL:  3

  XPASS: 1

  XFAIL: 2

 

Details:

 

 FAIL: 11_create_concurrent_pos

         [34] Failed to create domain

 

XFAIL: 02_network_local_ping_pos

         ping loopback failed for size 65507. ping eth0 failed for size
65507.

 

XFAIL: 11_network_domU_ping_pos

         Ping failed for size 65507.

 

 

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

 

 


[-- Attachment #1.2: Type: text/html, Size: 25300 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

Test results on Unisys ES7000 64x 256gb using xen-3.1.2 c/s 15502 - 2 old issues

[Krysan, Susan]

[-- Attachment #1.1: Type: text/plain, Size: 2281 bytes --]

Host:  Unisys ES7000/one, x86_64, 64 processors, 256gb memory

xen-3.1.2 changeset 15502 compiled with max_phys_cpus=64 and booted with
dom0_mem=4096M, numa=on, and xenheap_megabytes=64

 

OLD ISSUES:

-Bug #1037 - Shutdown of large domains takes a long time, during which
time dom0 is not interruptible (due to the synchronous tearing down of
the memory map); for example, shutdown of 180gb domu causes 5 min 15 sec
dom0 unresponsiveness  

-Bug #940 - Must specify hpet=disable kernel parameter to get 32-bit
SLES10 domVTs to boot; narrowed down to c/s 14436 but the patch
originator cannot recreate on his hardware

 

TESTS include running xm-test and also attempting to boot and run
programs in the following domUs and domVTs (running domains #s 3 through
10 simultaneously):

1)    32-processor 64-bit SLES10 domU with 248gb memory - run kernbench

2)    32-processor 64-bit SLES10 domVT with 248gb memory - run kernbench

3)    8-processor 64-bit SLES10 domU with 32gb memory - run kernbench

4)    8-processor 64-bit SLES10 domVT with 32gb memory - run kernbench

5)    8-processor 32-bit SLES10 domU with 32gb memory - run kernbench

6)    8-processor 32-bit PAE SLES10 domVT with 32gb memory (booted with
hpet=disable) - run kernbench

7)    8-processor 32-bit SLES10 domVT with 4gb memory (booted with
hpet=disable) - run kernbench

8)    8-processor 32-bit Windows 2003 Enterprise Edition domVT with 32gb
memory - run cpu intensive program

9)    8-processor 64-bit Windows Datacenter with 32gb memory - run cpu
intensive program

10)  2-processor 32-bit Windows XP domVT with 4gb memory - run cpu
intensive program

 

RESULTS:

All tests were successful, except for 2 old issues (see above).

 

Xm-test timing summary:

  Run Started : Mon, 26 Nov 2007 12:16:50 -0500

  Run Stoped  : Mon, 26 Nov 2007 13:10:25 -0500

Xm-test execution summary:

  PASS:  111

  FAIL:  2

  XPASS: 1

  XFAIL: 2

Details:

XFAIL: 02_network_local_ping_pos

         ping loopback failed for size 65507. ping eth0 failed for size
65507.

XFAIL: 11_network_domU_ping_pos

         Ping failed for size 65507.

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

 

 


[-- Attachment #1.2: Type: text/html, Size: 15887 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

Test results on Unisys ES7000 48x 166gb using unstable c/s 16471 - 2 old issues; 1 new but already being fixed

[Krysan, Susan]

[-- Attachment #1.1: Type: text/plain, Size: 2774 bytes --]

Host:  Unisys ES7000/one, x86_64, 48 processors, 166gb memory (host has
192gb, but truncates to 166gb)

unstable changeset 16471 compiled with max_phys_cpus=64 and booted with
dom0_mem=4096M, numa=on, and xenheap_megabytes=64

 

NEW ISSUE (fix on the way):

-four xm-test failures: ERROR Internal error: No extended context for
VCPU0

 

OLD ISSUES:

-Bug #1037 - Shutdown of large domains takes a long time, during which
time dom0 is not interruptible (due to the synchronous tearing down of
the memory map); for example, shutdown of 180gb domu causes 5 min 15 sec
dom0 unresponsiveness  

-Bug #940 - Must specify hpet=disable kernel parameter to get 32-bit
SLES10 domVTs to boot; narrowed down to c/s 14436 but the patch
originator cannot recreate on his hardware

 

TESTS include running xm-test and also attempting to boot and run
programs in the following domUs and domVTs (running domains #s 1 through
8 simultaneously):

1)    8-processor 64-bit SLES10 domU with 32gb memory - run kernbench

2)    8-processor 64-bit SLES10 domVT with 32gb memory - run kernbench

3)    8-processor 32-bit SLES10 domU with 16gb memory - run kernbench

4)    8-processor 32-bit PAE SLES10 domVT with 16gb memory (booted with
hpet=disable) - run kernbench

5)    8-processor 32-bit SLES10 domVT with 4gb memory (booted with
hpet=disable) - run kernbench

6)    8-processor 32-bit Windows 2003 Enterprise Edition domVT with 32gb
memory - run cpu intensive program

7)    8-processor 64-bit Windows Datacenter with 32gb memory - run cpu
intensive program

8)    2-processor 32-bit Windows XP domVT with 4gb memory - run cpu
intensive program

9)    32-processor 64-bit SLES10 domU with 160gb memory - run kernbench

10)  32-processor 64-bit SLES10 domVT with 160gb memory - run kernbench

 

RESULTS:

All tests were successful, except for 2 old issues (see above).

 

Xm-test timing summary:

  Run Started : Tue, 27 Nov 2007 15:16:42 -0500

  Run Stoped  : Tue, 27 Nov 2007 16:04:37 -0500

Xm-test execution summary:

  PASS:  106

  FAIL:  6

  XPASS: 1

  XFAIL: 2

 

Details:

 

 FAIL: 01_migrate_localhost_pos

         xm migrate returned invalid 256 != 0

 

XFAIL: 02_network_local_ping_pos

         ping loopback failed for size 65507. ping eth0 failed for size
65507.

 

XFAIL: 11_network_domU_ping_pos

         Ping failed for size 65507.

 

 FAIL: 01_restore_basic_pos

         save command exited 256 != 0

 

FAIL: 04_restore_withdevices_pos

         xm save exited with 256 != 0

 

 FAIL: 01_save_basic_pos

         save command exited 256 != 0

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

 

 


[-- Attachment #1.2: Type: text/html, Size: 14628 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

Test results on Unisys ES7000 64x 256gb using unstable c/s 16549 - 2 new issues (fix on the way for one); 2 old issues

[Krysan, Susan]

[-- Attachment #1.1: Type: text/plain, Size: 3091 bytes --]

Host:  Unisys ES7000/one, x86_64, 64 processors, 256gb memory

unstable changeset 16549 compiled with max_phys_cpus=64 and booted with
dom0_mem=4096M, numa=on, and xenheap_megabytes=64

 

NEW ISSUE (fix on the way):

-xm-test 02_block_attach fails with device /dev/ram1 does not exist

-xm-test 10_security-acm fails with KeyError 'access_control' (fix on
the way - Stefan Berger)

 

OLD ISSUES:

-Bug #1037 - Shutdown of large domains takes a long time, during which
time xend is unresponsive (due to the synchronous tearing down of the
memory map); for example, shutdown of 180gb domu causes xend commands to
hang for 5 min 15 sec

-Bug #940 - Must specify hpet=disable kernel parameter to get 32-bit
SLES10 domVTs to boot; narrowed down to c/s 14436 but the patch
originator cannot recreate on his hardware

 

TESTS include running xm-test and also booting and running programs in
the following domUs and domVTs (running domains #s 1 through 8
simultaneously):

1)    8-processor 64-bit SLES10 domU with 32gb memory - run kernbench

2)    8-processor 64-bit SLES10 domVT with 32gb memory - run kernbench

3)    8-processor 32-bit SLES10 domU with 32gb memory

4)    8-processor 32-bit PAE SLES10 domVT with 32gb memory (booted with
hpet=disable) 

5)    8-processor 32-bit SLES10 domVT with 4gb memory (booted with
hpet=disable) - run kernbench

6)    8-processor 32-bit Windows 2003 Enterprise Edition domVT with 32gb
memory - run cpu intensive program

7)    8-processor 64-bit Windows Datacenter with 32gb memory - run cpu
intensive program

8)    2-processor 32-bit Windows XP domVT with 4gb memory - run cpu
intensive program

9)    32-processor 64-bit SLES10 domU with 240gb memory

10)  32-processor 64-bit SLES10 domVT with 240gb memory

 

RESULTS:

All tests were successful, except for 2 old issues (see above).

 

02_block_attach fails with:

[dom0] Running `xm block-attach
02_block_attach_file_device_pos-1196993485 file:/dev/ram1 xvda1 w'

Error: Device 51713 (vbd) could not be connected. /dev/ram1 does not
exist.

Usage: xm block-attach <Domain> <BackDev> <FrontDev> <Mode> [BackDomain]

 

10_security-acm fails with:

  File "./10_security-acm_pol_update.test", line 40, in ?

    del domain_ul.config.opts['access_control']

KeyError: 'access_control'

 

Xm-test timing summary:

  Run Started : Thu, 06 Dec 2007 21:11:12 -0500

  Run Stoped  : Thu, 06 Dec 2007 22:01:10 -0500

Xm-test execution summary:

  PASS:  109

  FAIL:  4

  XPASS: 1

  XFAIL: 2

 

 

Details:

 

 FAIL: 02_block_attach_file_device_pos

         xm block-attach returned invalid 256 != 0

 

XFAIL: 02_network_local_ping_pos

         ping loopback failed for size 65507. ping eth0 failed for size
65507.

 

XFAIL: 11_network_domU_ping_pos

         Ping failed for size 1 48 64 512 1440 65507.

 

 FAIL: 10_security-acm_pol_update

        Unknown reason

 

Thanks,

Sue Krysan

Linux Systems Group

Unisys Corporation

 

 

 


[-- Attachment #1.2: Type: text/html, Size: 15928 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel