The Ceph wiki?

The Ceph wiki?

[Daniel Friesen]

I noticed the Ceph wiki is fairly out of date, and wrought with the 
usual spam (well, having .pdf enabled entitles a type of spam I haven't 
seen before).
Which felt a little concerning, since the wiki appears to be the primary 
source of documentation, and Ceph looks like an active project with a 
lot of potential.

Could the Ceph group use my help maintaining the wiki?

-- 
~Daniel Friesen (Dantman, Nadir-Seen-Fire) [http://daniel.friesen.name]

Re: The Ceph wiki?

[Sage Weil]

On Sun, 13 Feb 2011, Daniel Friesen wrote:
> I noticed the Ceph wiki is fairly out of date, and wrought with the usual spam
> (well, having .pdf enabled entitles a type of spam I haven't seen before).
> Which felt a little concerning, since the wiki appears to be the primary
> source of documentation, and Ceph looks like an active project with a lot of
> potential.
> 
> Could the Ceph group use my help maintaining the wiki?

Absolutely!  Let me know what (if any) privs you need to work with it, or 
what ought to be fixed up with the configuration.

sage

Re: The Ceph wiki?

[Daniel Friesen]

On 11-02-13 10:30 AM, Sage Weil wrote:
> On Sun, 13 Feb 2011, Daniel Friesen wrote:
>> I noticed the Ceph wiki is fairly out of date, and wrought with the usual spam
>> (well, having .pdf enabled entitles a type of spam I haven't seen before).
>> Which felt a little concerning, since the wiki appears to be the primary
>> source of documentation, and Ceph looks like an active project with a lot of
>> potential.
>>
>> Could the Ceph group use my help maintaining the wiki?
> Absolutely!  Let me know what (if any) privs you need to work with it, or
> what ought to be fixed up with the configuration.
>
> sage
- Upgrading and installation of various anti-spam extensions still need 
commands to be run, so wherever it is I'll need shell access...
-- and a mysql account with allprivs on the database the wiki is on.
-- I prefer to maintain MW by checking it out from svn anyways (the 
REL1_## branches).
- I'll probably also setup the usual cron job I use (job queue in cron, 
sitemap generation, ...) so I'll need to be able to create one of those.
- I'm undecided on whether I should setup a 404 image handler since 
there aren't many images being used (a handler speeds up page loads by 
not making them dependent on the completion of thumbnail generation).

Whether you want me in the servers, or want me to host it on one of mine 
is up to you. I'm already hosting the CommonJS wiki and website.

-- 
~Daniel Friesen (Dantman, Nadir-Seen-Fire) [http://daniel.friesen.name]

how long client will take when recover to ok,after ceph down.

[Longguang Yue]

When client lose its connection with ceph, then ls /mnt/ceph, it will
hang all the time
Remount does not work.
Thanks

question about socket conection among osd,mds,mon

[Longguang Yue]

1.Netstat info:
10.10.1.100:6800		<--- 10.10.1.106:34585, 10.10.1.98:46625
10.10.1.100:6801		<--- 10.10.1.106:59284, 10.10.1.98:35101
When osd need 2(more) listen sockets? Why  others osd connect as many
times as its listen socket?


2.how to deal with blacklist osds
osd e35: 3 osds: 3 up, 3 in -- 2 blacklisted MDSes

according to 1 case, how to deal with blacklist.
If I need  run  'ceph osd blacklist rm  10.10.1.106:34585 &&
10.10.1.98:46625 ' ??????

I think just only one osd listen socket exists!!!!!!

Re: how long client will take when recover to ok,after ceph down.

[Colin McCabe]

Hi Longguang,

Basically, if *all* remote servers become inaccessible, you have two
bad choices:

1) Wait for the remote servers to become accessible.
This is Ceph's current behavior. This is also NFS' behavior in its
default "hard mount" mode.

2) Throw away data and metadata, but keep running
This is NFS' behavior if you have done a "soft mount."

Neither one of these choices is really good. That's why Ceph's focus
is on keeping the filesystem running even if a few nodes go down. We
have talked about implementing "soft mount" semantics for Ceph, but I
don't think it's been done yet (Sage, correct me if I'm wrong here?)
Also, you could get soft mount semantics by using libceph to access
the filesystem rather than the kernel client. Userspace programs usnig
libceph can always be killed in mid-write or read.

A lot of system administrators don't like soft mount semantics though.
Think carefully if that's really what you want.

cheers,
Colin


On Sun, Feb 13, 2011 at 5:21 PM, Longguang Yue
<longguang_yue@tcloudcomputing.com> wrote:
> When client lose its connection with ceph, then ls /mnt/ceph, it will
> hang all the time
> Remount does not work.
> Thanks
> --
> To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>
--
To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: question about socket conection among osd,mds,mon

[Colin McCabe]

Hi Longguang,

The blacklist stops MDSes from talking to OSDs. You cannot blacklist
an OSD itself.

What is it that you're trying to do?

sincerely,
Colin


On Sun, Feb 13, 2011 at 7:13 PM, Longguang Yue
<longguang_yue@tcloudcomputing.com> wrote:
> 1.Netstat info:
> 10.10.1.100:6800                <--- 10.10.1.106:34585, 10.10.1.98:46625
> 10.10.1.100:6801                <--- 10.10.1.106:59284, 10.10.1.98:35101
> When osd need 2(more) listen sockets? Why  others osd connect as many
> times as its listen socket?
>
>
> 2.how to deal with blacklist osds
> osd e35: 3 osds: 3 up, 3 in -- 2 blacklisted MDSes
>
> according to 1 case, how to deal with blacklist.
> If I need  run  'ceph osd blacklist rm  10.10.1.106:34585 &&
> 10.10.1.98:46625 ' ??????
>
> I think just only one osd listen socket exists!!!!!!
> --
> To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>
--
To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

答复: question about socket conection among osd,mds,mon

[Longguang Yue]

!!!!    osd e35: 3 osds: 3 up, 3 in -- 2 blacklisted MDSes

ceph can recover blacklisted osds by itself, but I want to recover blacklist osds as soon as possible.

How to use the command,		'ceph osd blacklist rm host:port' ????

thanks

-----邮件原件-----
发件人: rarecactus@gmail.com [mailto:rarecactus@gmail.com] 代表 Colin McCabe
发送时间: 2011年2月14日 18:21
收件人: Longguang Yue
抄送: Sage Weil; ceph-devel@vger.kernel.org
主题: Re: question about socket conection among osd,mds,mon

Hi Longguang,

The blacklist stops MDSes from talking to OSDs. You cannot blacklist
an OSD itself.

What is it that you're trying to do?

sincerely,
Colin


On Sun, Feb 13, 2011 at 7:13 PM, Longguang Yue
<longguang_yue@tcloudcomputing.com> wrote:
> 1.Netstat info:
> 10.10.1.100:6800                <--- 10.10.1.106:34585, 10.10.1.98:46625
> 10.10.1.100:6801                <--- 10.10.1.106:59284, 10.10.1.98:35101
> When osd need 2(more) listen sockets? Why  others osd connect as many
> times as its listen socket?
>
>
> 2.how to deal with blacklist osds
> osd e35: 3 osds: 3 up, 3 in -- 2 blacklisted MDSes
>
> according to 1 case, how to deal with blacklist.
> If I need  run  'ceph osd blacklist rm  10.10.1.106:34585 &&
> 10.10.1.98:46625 ' ??????
>
> I think just only one osd listen socket exists!!!!!!
> --
> To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>

Re: question about socket conection among osd,mds,mon

[Gregory Farnum]

On Monday, February 14, 2011 at 5:37 AM, Longguang Yue wrote: 
> !!!! osd e35: 3 osds: 3 up, 3 in -- 2 blacklisted MDSes
> 
> ceph can recover blacklisted osds by itself, but I want to recover blacklist osds as soon as possible.
> 
> How to use the command, 'ceph osd blacklist rm host:port' ????
> 
> thanks
None of your OSDs have been blacklisted. 2 of your MDSes have been, but that means that they either went down or were so laggy as to be useless. If they were laggy, then they will eventually find out and restart themselves, at which point they will be identified as a different MDS and won't be blacklisted.
This doesn't mean there's anything wrong with your OSDs, though!
-Greg 

> -----邮件原件-----
> 发件人: rarecactus@gmail.com [mailto:rarecactus@gmail.com] 代表 Colin McCabe
> 发送时间: 2011年2月14日 18:21
> 收件人: Longguang Yue
> 抄送: Sage Weil; ceph-devel@vger.kernel.org
> 主题: Re: question about socket conection among osd,mds,mon
> 
> Hi Longguang,
> 
> The blacklist stops MDSes from talking to OSDs. You cannot blacklist
> an OSD itself.
> 
> What is it that you're trying to do?
> 
> sincerely,
> Colin
> 
> 
> On Sun, Feb 13, 2011 at 7:13 PM, Longguang Yue
> <longguang_yue@tcloudcomputing.com> wrote:
> > 1.Netstat info:
> > 10.10.1.100:6800 <--- 10.10.1.106:34585, 10.10.1.98:46625
> > 10.10.1.100:6801 <--- 10.10.1.106:59284, 10.10.1.98:35101
> > When osd need 2(more) listen sockets? Why others osd connect as many
> > times as its listen socket?
> > 
> > 
> > 2.how to deal with blacklist osds
> > osd e35: 3 osds: 3 up, 3 in -- 2 blacklisted MDSes
> > 
> > according to 1 case, how to deal with blacklist.
> > If I need run 'ceph osd blacklist rm 10.10.1.106:34585 &&
> > 10.10.1.98:46625 ' ??????
> > 
> > I think just only one osd listen socket exists!!!!!!
> > --
> > To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
> > the body of a message to majordomo@vger.kernel.org
> > More majordomo info at http://vger.kernel.org/majordomo-info.html
> ??{.n?+???????+%??lzwm??b?맲??r??yǩ?ׯzX????ܨ}???Ơz?&j:+v???????zZ+??+zf???h???~????i???z??w?????????&?)ߢf?

--
To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html