All of lore.kernel.org
 help / color / mirror / Atom feed
* [LTP] [PATCH 1/2] tst_net.sh: Add API function to check netns use
@ 2021-02-05 14:20 Petr Vorel
  2021-02-05 14:20 ` [LTP] [PATCH 2/2] if-mtu-change.sh: Lower CHANGE_INTERVAL for netns Petr Vorel
  0 siblings, 1 reply; 6+ messages in thread
From: Petr Vorel @ 2021-02-05 14:20 UTC (permalink / raw)
  To: ltp

Signed-off-by: Petr Vorel <pvorel@suse.cz>
---
Hi Alexey,

or do we want to allow using TST_USE_NETNS?

Kind regards,
Petr

 testcases/lib/tst_net.sh | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/testcases/lib/tst_net.sh b/testcases/lib/tst_net.sh
index 1ddef2fea..d4633c509 100644
--- a/testcases/lib/tst_net.sh
+++ b/testcases/lib/tst_net.sh
@@ -131,6 +131,12 @@ init_ltp_netspace()
 	tst_restore_ipaddr rhost
 }
 
+# return 0: use ssh, 1: use netns
+tst_net_use_netns()
+{
+	[ -n "$TST_USE_NETNS" ]
+}
+
 # Run command on remote host.
 # tst_rhost_run -c CMD [-b] [-s] [-u USER]
 # Options:
-- 
2.30.0


^ permalink raw reply related	[flat|nested] 6+ messages in thread

* [LTP] [PATCH 2/2] if-mtu-change.sh: Lower CHANGE_INTERVAL for netns
  2021-02-05 14:20 [LTP] [PATCH 1/2] tst_net.sh: Add API function to check netns use Petr Vorel
@ 2021-02-05 14:20 ` Petr Vorel
  2021-02-08 12:49   ` Alexey Kodanev
  0 siblings, 1 reply; 6+ messages in thread
From: Petr Vorel @ 2021-02-05 14:20 UTC (permalink / raw)
  To: ltp

to make testing faster.

NOTE: the default sleep 5s is needed for ssh based testing to prevent
link-flap errors on the switch. It can be lower down by user.

Signed-off-by: Petr Vorel <pvorel@suse.cz>
---
 .../network/stress/interface/if-mtu-change.sh      | 14 ++++++++++----
 1 file changed, 10 insertions(+), 4 deletions(-)

diff --git a/testcases/network/stress/interface/if-mtu-change.sh b/testcases/network/stress/interface/if-mtu-change.sh
index d2816606b..49ac5e3fc 100755
--- a/testcases/network/stress/interface/if-mtu-change.sh
+++ b/testcases/network/stress/interface/if-mtu-change.sh
@@ -1,6 +1,6 @@
 #!/bin/sh
 # SPDX-License-Identifier: GPL-2.0-or-later
-# Copyright (c) 2017-2019 Petr Vorel <pvorel@suse.cz>
+# Copyright (c) 2017-2021 Petr Vorel <pvorel@suse.cz>
 # Copyright (c) 2015-2017 Oracle and/or its affiliates. All Rights Reserved.
 # Copyright (c) International Business Machines  Corp., 2005
 # Author: Mitsuru Chinen <mitch@jp.ibm.com>
@@ -12,9 +12,13 @@ TST_CLEANUP="do_cleanup"
 
 # The interval of the mtu change [second]
 CHANGE_INTERVAL=${CHANGE_INTERVAL:-5}
-
 TST_TIMEOUT=$(((CHANGE_INTERVAL + 30) * MTU_CHANGE_TIMES))
 
+if tst_net_use_netns; then
+	CHANGE_INTERVAL="100ms"
+	TST_TIMEOUT=$((30 * MTU_CHANGE_TIMES))
+fi
+
 # The array of the value which MTU is changed into sequentially
 # 552 - net.ipv4.route.min_pmtu
 CHANGE_VALUES="784 1142 552 1500 552 1500 552 748 552 1142 1500"
@@ -65,6 +69,7 @@ find_ipv4_max_packet_size()
 
 do_setup()
 {
+
 	[ "$TST_IPV6" ] && CHANGE_VALUES=$CHANGE6_VALUES
 	if_setup
 	saved_mtu="$(cat /sys/class/net/$(tst_iface)/mtu)"
@@ -83,9 +88,10 @@ do_cleanup()
 test_body()
 {
 	local cmd="$CMD"
+	local msg="'$cmd' changes MTU $MTU_CHANGE_TIMES times every $CHANGE_INTERVAL"
 
-	tst_res TINFO "'$cmd' changes MTU $MTU_CHANGE_TIMES times" \
-	               "every $CHANGE_INTERVAL seconds"
+	tst_is_int $CHANGE_INTERVAL && msg="$msg seconds"
+	tst_res TINFO "$msg"
 
 	mtu_array_len=$(echo $CHANGE_VALUES | wc -w)
 	local cnt=0
-- 
2.30.0


^ permalink raw reply related	[flat|nested] 6+ messages in thread

* [LTP] [PATCH 2/2] if-mtu-change.sh: Lower CHANGE_INTERVAL for netns
  2021-02-05 14:20 ` [LTP] [PATCH 2/2] if-mtu-change.sh: Lower CHANGE_INTERVAL for netns Petr Vorel
@ 2021-02-08 12:49   ` Alexey Kodanev
  2021-02-08 16:04     ` Petr Vorel
  0 siblings, 1 reply; 6+ messages in thread
From: Alexey Kodanev @ 2021-02-08 12:49 UTC (permalink / raw)
  To: ltp

On 05.02.2021 17:20, Petr Vorel wrote:
> to make testing faster.
> 
> NOTE: the default sleep 5s is needed for ssh based testing to prevent
> link-flap errors on the switch. It can be lower down by user.
> 
> Signed-off-by: Petr Vorel <pvorel@suse.cz>
> ---
>  .../network/stress/interface/if-mtu-change.sh      | 14 ++++++++++----
>  1 file changed, 10 insertions(+), 4 deletions(-)
> 
> diff --git a/testcases/network/stress/interface/if-mtu-change.sh b/testcases/network/stress/interface/if-mtu-change.sh
> index d2816606b..49ac5e3fc 100755
> --- a/testcases/network/stress/interface/if-mtu-change.sh
> +++ b/testcases/network/stress/interface/if-mtu-change.sh
> @@ -1,6 +1,6 @@
>  #!/bin/sh
>  # SPDX-License-Identifier: GPL-2.0-or-later
> -# Copyright (c) 2017-2019 Petr Vorel <pvorel@suse.cz>
> +# Copyright (c) 2017-2021 Petr Vorel <pvorel@suse.cz>
>  # Copyright (c) 2015-2017 Oracle and/or its affiliates. All Rights Reserved.
>  # Copyright (c) International Business Machines  Corp., 2005
>  # Author: Mitsuru Chinen <mitch@jp.ibm.com>
> @@ -12,9 +12,13 @@ TST_CLEANUP="do_cleanup"
>  
>  # The interval of the mtu change [second]
>  CHANGE_INTERVAL=${CHANGE_INTERVAL:-5}
> -
>  TST_TIMEOUT=$(((CHANGE_INTERVAL + 30) * MTU_CHANGE_TIMES))
>  
> +if tst_net_use_netns; then
> +	CHANGE_INTERVAL="100ms"

Hi Petr,

What about adding the option to change this value too for netns,
for example like this:

TST_TIMEOUT=1
if tst_net_use_netns; then
    CHANGE_INTERVAL=${CHANGE_INTERVAL:-100ms}
else
    CHANGE_INTERVAL=${CHANGE_INTERVAL:-5}
    tst_is_int $CHANGE_INTERVAL && TST_TIMEOUT=$CHANGE_INTERVAL
fi

TST_TIMEOUT=$(((TST_TIMEOUT + 30) * MTU_CHANGE_TIMES))


> +	TST_TIMEOUT=$((30 * MTU_CHANGE_TIMES))
> +fi
> +
>  # The array of the value which MTU is changed into sequentially
>  # 552 - net.ipv4.route.min_pmtu
>  CHANGE_VALUES="784 1142 552 1500 552 1500 552 748 552 1142 1500"
> @@ -65,6 +69,7 @@ find_ipv4_max_packet_size()
>  
>  do_setup()
>  {
> +

empty new line?

>  	[ "$TST_IPV6" ] && CHANGE_VALUES=$CHANGE6_VALUES
>  	if_setup
>  	saved_mtu="$(cat /sys/class/net/$(tst_iface)/mtu)"
> @@ -83,9 +88,10 @@ do_cleanup()
>  test_body()
>  {
>  	local cmd="$CMD"
> +	local msg="'$cmd' changes MTU $MTU_CHANGE_TIMES times every $CHANGE_INTERVAL"
>  
> -	tst_res TINFO "'$cmd' changes MTU $MTU_CHANGE_TIMES times" \
> -	               "every $CHANGE_INTERVAL seconds"
> +	tst_is_int $CHANGE_INTERVAL && msg="$msg seconds"
> +	tst_res TINFO "$msg"
>  
>  	mtu_array_len=$(echo $CHANGE_VALUES | wc -w)
>  	local cnt=0
>

^ permalink raw reply	[flat|nested] 6+ messages in thread

* [LTP] [PATCH 2/2] if-mtu-change.sh: Lower CHANGE_INTERVAL for netns
  2021-02-08 12:49   ` Alexey Kodanev
@ 2021-02-08 16:04     ` Petr Vorel
  2021-02-09 12:45       ` Alexey Kodanev
  0 siblings, 1 reply; 6+ messages in thread
From: Petr Vorel @ 2021-02-08 16:04 UTC (permalink / raw)
  To: ltp

Hi Alexey,

> >  # The interval of the mtu change [second]
> >  CHANGE_INTERVAL=${CHANGE_INTERVAL:-5}
> > -
> >  TST_TIMEOUT=$(((CHANGE_INTERVAL + 30) * MTU_CHANGE_TIMES))

> > +if tst_net_use_netns; then
> > +	CHANGE_INTERVAL="100ms"

> Hi Petr,

> What about adding the option to change this value too for netns,
> for example like this:

> TST_TIMEOUT=1
> if tst_net_use_netns; then
>     CHANGE_INTERVAL=${CHANGE_INTERVAL:-100ms}
> else
>     CHANGE_INTERVAL=${CHANGE_INTERVAL:-5}
>     tst_is_int $CHANGE_INTERVAL && TST_TIMEOUT=$CHANGE_INTERVAL
> fi

> TST_TIMEOUT=$(((TST_TIMEOUT + 30) * MTU_CHANGE_TIMES))
That's indeed much better!

How about allow to overwrite CHANGE_INTERVAL even for netns?
i.e move tst_is_int $CHANGE_INTERVAL && TST_TIMEOUT=$CHANGE_INTERVAL
after the test (although it's unlikely anybody would use sec value for netns
and ms value for ssh):

# CHANGE_INTERVAL: The interval of the mtu change
TST_TIMEOUT=1
if tst_net_use_netns; then
    CHANGE_INTERVAL=${CHANGE_INTERVAL:-100ms}
else
    CHANGE_INTERVAL=${CHANGE_INTERVAL:-5}
fi
tst_is_int $CHANGE_INTERVAL && TST_TIMEOUT=$CHANGE_INTERVAL
TST_TIMEOUT=$(((TST_TIMEOUT + 30) * MTU_CHANGE_TIMES))

> > +	TST_TIMEOUT=$((30 * MTU_CHANGE_TIMES))
> > +fi
> > +
> >  # The array of the value which MTU is changed into sequentially
> >  # 552 - net.ipv4.route.min_pmtu
> >  CHANGE_VALUES="784 1142 552 1500 552 1500 552 748 552 1142 1500"
> > @@ -65,6 +69,7 @@ find_ipv4_max_packet_size()

> >  do_setup()
> >  {
> > +

> empty new line?
Thanks!

Kind regards,
Petr

^ permalink raw reply	[flat|nested] 6+ messages in thread

* [LTP] [PATCH 2/2] if-mtu-change.sh: Lower CHANGE_INTERVAL for netns
  2021-02-08 16:04     ` Petr Vorel
@ 2021-02-09 12:45       ` Alexey Kodanev
  2021-02-09 14:12         ` Petr Vorel
  0 siblings, 1 reply; 6+ messages in thread
From: Alexey Kodanev @ 2021-02-09 12:45 UTC (permalink / raw)
  To: ltp

On 08.02.2021 19:04, Petr Vorel wrote:
> How about allow to overwrite CHANGE_INTERVAL even for netns?
> i.e move tst_is_int $CHANGE_INTERVAL && TST_TIMEOUT=$CHANGE_INTERVAL
> after the test (although it's unlikely anybody would use sec value for netns
> and ms value for ssh):
> 
> # CHANGE_INTERVAL: The interval of the mtu change
> TST_TIMEOUT=1
> if tst_net_use_netns; then
>     CHANGE_INTERVAL=${CHANGE_INTERVAL:-100ms}
> else
>     CHANGE_INTERVAL=${CHANGE_INTERVAL:-5}
> fi
> tst_is_int $CHANGE_INTERVAL && TST_TIMEOUT=$CHANGE_INTERVAL
> TST_TIMEOUT=$(((TST_TIMEOUT + 30) * MTU_CHANGE_TIMES))
> 

Yeah, looks good.

^ permalink raw reply	[flat|nested] 6+ messages in thread

* [LTP] [PATCH 2/2] if-mtu-change.sh: Lower CHANGE_INTERVAL for netns
  2021-02-09 12:45       ` Alexey Kodanev
@ 2021-02-09 14:12         ` Petr Vorel
  0 siblings, 0 replies; 6+ messages in thread
From: Petr Vorel @ 2021-02-09 14:12 UTC (permalink / raw)
  To: ltp

Hi Alexey,

> On 08.02.2021 19:04, Petr Vorel wrote:
> > How about allow to overwrite CHANGE_INTERVAL even for netns?
> > i.e move tst_is_int $CHANGE_INTERVAL && TST_TIMEOUT=$CHANGE_INTERVAL
> > after the test (although it's unlikely anybody would use sec value for netns
> > and ms value for ssh):

> > # CHANGE_INTERVAL: The interval of the mtu change
> > TST_TIMEOUT=1
> > if tst_net_use_netns; then
> >     CHANGE_INTERVAL=${CHANGE_INTERVAL:-100ms}
> > else
> >     CHANGE_INTERVAL=${CHANGE_INTERVAL:-5}
> > fi
> > tst_is_int $CHANGE_INTERVAL && TST_TIMEOUT=$CHANGE_INTERVAL
> > TST_TIMEOUT=$(((TST_TIMEOUT + 30) * MTU_CHANGE_TIMES))


> Yeah, looks good.
Thanks for your ack, merged!

Kind regards,
Petr

^ permalink raw reply	[flat|nested] 6+ messages in thread

end of thread, other threads:[~2021-02-09 14:12 UTC | newest]

Thread overview: 6+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-02-05 14:20 [LTP] [PATCH 1/2] tst_net.sh: Add API function to check netns use Petr Vorel
2021-02-05 14:20 ` [LTP] [PATCH 2/2] if-mtu-change.sh: Lower CHANGE_INTERVAL for netns Petr Vorel
2021-02-08 12:49   ` Alexey Kodanev
2021-02-08 16:04     ` Petr Vorel
2021-02-09 12:45       ` Alexey Kodanev
2021-02-09 14:12         ` Petr Vorel

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.