hi, <br>Do you more information about this bug ?<br>I do not understand how afr works, <br>with my initial configuration, if i change ip of network card ( from 10.98.98.2 => 10.98.98.4 ) on server B during test, <br>on client and server (A ,C ) 'ls' works after some timeout, but some program seems to be block all system (<br>
if i run my own program or qemu for example) 'ls' does not respond anymore, and if i rechange from 10.98.98.4 => 10.98.98.2 ) then all become ok again. <br><br>Regards,<br>Nicolas Prochazka<br><br><br><div class="gmail_quote">
2009/1/14 Krishna Srinivas <span dir="ltr"><<a href="mailto:krishna@zresearch.com">krishna@zresearch.com</a>></span><br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
Nicolas,<br>
<br>
It might be a bug. Let me try to reproduce the problem here and get back to you.<br>
<br>
Krishna<br>
<br>
On Wed, Jan 14, 2009 at 6:59 PM, nicolas prochazka<br>
<div><div></div><div class="Wj3C7c"><<a href="mailto:prochazka.nicolas@gmail.com">prochazka.nicolas@gmail.com</a>> wrote:<br>
> hello again,<br>
> To finish with this issue and information I can send you :<br>
> If i stop glusterfsd ( on server B) before to stop this server ( hard<br>
> poweroff by pressed on/off ) , the problem does not occur. If i hard<br>
> poweroff without stop gluster ( a real crash ) problem occur .<br>
> Regards<br>
> Nicolas Prochazka.<br>
><br>
> 2009/1/14 nicolas prochazka <<a href="mailto:prochazka.nicolas@gmail.com">prochazka.nicolas@gmail.com</a>><br>
>><br>
>> hi again,<br>
>> I continue my tests and :<br>
>> In my case, if one file is open on gluster mount during stop of one afr<br>
>> server,<br>
>> gluster mount can not be acces ( gap ? ) in this server. All other client<br>
>> ( C for example) which not opening file during stop, isn't affect, i can do<br>
>> a ls or open after transport timeout time.<br>
>> If i kill the process that's use this file, then i can using gluster mount<br>
>> point without problem.<br>
>><br>
>> Regards,<br>
>> Nicolas Prochazka.<br>
>><br>
>> 2009/1/12 nicolas prochazka <<a href="mailto:prochazka.nicolas@gmail.com">prochazka.nicolas@gmail.com</a>><br>
>>><br>
>>> for your attention,<br>
>>> it seems that's this problem occur only when files is open and use and<br>
>>> gluster mount point .<br>
>>> I use big files of computation ( ~ 10G) with in the most important part,<br>
>>> read. In this case problem occurs.<br>
>>> If i using only small files which create only some time, no problem<br>
>>> occur, gluster mount can use other afr server.<br>
>>><br>
>>> Regards,<br>
>>> Nicolas Prochazka<br>
>>><br>
>>><br>
>>><br>
>>> 2009/1/12 nicolas prochazka <<a href="mailto:prochazka.nicolas@gmail.com">prochazka.nicolas@gmail.com</a>><br>
>>>><br>
>>>> Hi,<br>
>>>> I'm tryning to set<br>
>>>> option transport-timeout 5<br>
>>>> in protocol/client<br>
>>>><br>
>>>> so a max of 10 seconds before restoring gluster in normal situation ?<br>
>>>> no success, i always in the same situation, a 'ls /mnt/gluster' not<br>
>>>> respond after > 10 mins<br>
>>>> I can not reuse glustermount exept kill glusterfs process.<br>
>>>><br>
>>>> Regards<br>
>>>> Nicolas Prochazka<br>
>>>><br>
>>>><br>
>>>><br>
>>>> 2009/1/12 Raghavendra G <<a href="mailto:raghavendra@zresearch.com">raghavendra@zresearch.com</a>><br>
>>>>><br>
>>>>> Hi Nicolas,<br>
>>>>><br>
>>>>> how much time did you wait before concluding the mount point to be not<br>
>>>>> working? afr waits for a maximum of (2 * transport-timeout) seconds before<br>
>>>>> returning sending reply to the application. Can you wait for some time and<br>
>>>>> check out is this the issue you are facing?<br>
>>>>><br>
>>>>> regards,<br>
>>>>><br>
>>>>> On Mon, Jan 12, 2009 at 7:49 PM, nicolas prochazka<br>
>>>>> <<a href="mailto:prochazka.nicolas@gmail.com">prochazka.nicolas@gmail.com</a>> wrote:<br>
>>>>>><br>
>>>>>> Hi.<br>
>>>>>> I've installed this model to test Gluster :<br>
>>>>>><br>
>>>>>> + 2 servers ( A B )<br>
>>>>>> - with glusterfsd server ( glusterfs--mainline--3.0--patch-842 )<br>
>>>>>> - with glusterfs client<br>
>>>>>> server conf file .<br>
>>>>>><br>
>>>>>> + 1 server C only client mode.<br>
>>>>>><br>
>>>>>> My issue :<br>
>>>>>> If C open big file in this client configuration and then i stop server<br>
>>>>>> A (or B )<br>
>>>>>> gluster mount point on server C seems to be block, i can not do 'ls<br>
>>>>>> -l' for example.<br>
>>>>>> Is a this thing is normal ? as C open his file on A or B , then it is<br>
>>>>>> blocking when server down ?<br>
>>>>>> I was thinking in client AFR, client can reopen file/block an other<br>
>>>>>> server , i'm wrong ?<br>
>>>>>> Should use HA translator ?<br>
>>>>>><br>
>>>>>> Regards,<br>
>>>>>> Nicolas Prochazka.<br>
>>>>>><br>
>>>>>><br>
>>>>>><br>
>>>>>><br>
>>>>>><br>
>>>>>> volume brickless<br>
>>>>>> type storage/posix<br>
>>>>>> option directory /mnt/disks/export<br>
>>>>>> end-volume<br>
>>>>>><br>
>>>>>> volume brick<br>
>>>>>> type features/posix-locks<br>
>>>>>> option mandatory on # enables mandatory locking on all files<br>
>>>>>> subvolumes brickless<br>
>>>>>> end-volume<br>
>>>>>><br>
>>>>>> volume server<br>
>>>>>> type protocol/server<br>
>>>>>> subvolumes brick<br>
>>>>>> option transport-type tcp<br>
>>>>>> option auth.addr.brick.allow 10.98.98.*<br>
>>>>>> end-volume<br>
>>>>>> ---------------------------<br>
>>>>>><br>
>>>>>> client config<br>
>>>>>> volume brick_10.98.98.1<br>
>>>>>> type protocol/client<br>
>>>>>> option transport-type tcp/client<br>
>>>>>> option remote-host 10.98.98.1<br>
>>>>>> option remote-subvolume brick<br>
>>>>>> end-volume<br>
>>>>>><br>
>>>>>> volume brick_10.98.98.2<br>
>>>>>> type protocol/client<br>
>>>>>> option transport-type tcp/client<br>
>>>>>> option remote-host 10.98.98.2<br>
>>>>>> option remote-subvolume brick<br>
>>>>>> end-volume<br>
>>>>>><br>
>>>>>> volume last<br>
>>>>>> type cluster/replicate<br>
>>>>>> subvolumes brick_10.98.98.1 brick_10.98.98.2<br>
>>>>>> end-volume<br>
>>>>>><br>
>>>>>> volume iothreads<br>
>>>>>> type performance/io-threads<br>
>>>>>> option thread-count 2<br>
>>>>>> option cache-size 32MB<br>
>>>>>> subvolumes last<br>
>>>>>> end-volume<br>
>>>>>><br>
>>>>>> volume io-cache<br>
>>>>>> type performance/io-cache<br>
>>>>>> option cache-size 1024MB # default is 32MB<br>
>>>>>> option page-size 1MB #128KB is default option<br>
>>>>>> option force-revalidate-timeout 2 # default is 1<br>
>>>>>> subvolumes iothreads<br>
>>>>>> end-volume<br>
>>>>>><br>
>>>>>> volume writebehind<br>
>>>>>> type performance/write-behind<br>
>>>>>> option aggregate-size 256KB # default is 0bytes<br>
>>>>>> option window-size 3MB<br>
>>>>>> option flush-behind on # default is 'off'<br>
>>>>>> subvolumes io-cache<br>
>>>>>> end-volume<br>
>>>>>><br>
>>>>>><br>
>>>>>> _______________________________________________<br>
>>>>>> Gluster-devel mailing list<br>
>>>>>> <a href="mailto:Gluster-devel@nongnu.org">Gluster-devel@nongnu.org</a><br>
>>>>>> <a href="http://lists.nongnu.org/mailman/listinfo/gluster-devel" target="_blank">http://lists.nongnu.org/mailman/listinfo/gluster-devel</a><br>
>>>>>><br>
>>>>><br>
>>>>><br>
>>>>><br>
>>>>> --<br>
>>>>> Raghavendra G<br>
>>>>><br>
>>>><br>
>>><br>
>><br>
><br>
><br>
> _______________________________________________<br>
> Gluster-devel mailing list<br>
> <a href="mailto:Gluster-devel@nongnu.org">Gluster-devel@nongnu.org</a><br>
> <a href="http://lists.nongnu.org/mailman/listinfo/gluster-devel" target="_blank">http://lists.nongnu.org/mailman/listinfo/gluster-devel</a><br>
><br>
><br>
</div></div></blockquote></div><br>