[Gluster-users] when one of the server down , the delay too long

[Mohit Anchlia]

It will be great if developers can jump in here.

What behaviour are you seeing and what value have you set for ping
timeout. Are all the requests slow or only first few?

On Sun, Apr 10, 2011 at 1:32 AM,  <lihang at netopcomputing.com> wrote:
> On Sat, 9 Apr 2011 17:28:36 -0700, Mohit Anchlia wrote:
>>
>> Try network.ping-timeout in volume options and see if it helps. Post
>> your results after making that change.
>>
>>
>>
>> http://europe.gluster.org/community/documentation/index.php/Gluster_3.1:_Setting_Volume_Options
>>
>> On Sat, Apr 9, 2011 at 4:27 PM, Mohit Anchlia <mohitanchlia at gmail.com>
>> wrote:
>>>
>>> I think there is a socket timeout you might have to set. Can someone
>>> from developers pitch in?
>>>
>>> On Fri, Apr 8, 2011 at 10:21 PM, lihang <lihang at netopcomputing.com>
>>> wrote:
>>>>
>>>> I have two GlusterFS server and the volume status is replica . The add
>>>> of them are server01:192.168.1.10 and server02:192.168.1.11 . The client
>>>> mount the server01'vol and I can use the GLusterFS usually.
>>>>    Now, l am reading the file on the glusterfs volume usually, the
>>>> server02's interface down suddenly ， and the client is also down. It will
>>>> resume after a delay(about 10s),I think it's too long. It will make some
>>>> errors in the application ,please help me to the problem , thank you.
>>>> 发送自 HTC
>>>>
>>>>
>>>> _______________________________________________
>>>> Gluster-users mailing list
>>>> Gluster-users at gluster.org
>>>> http://gluster.org/cgi-bin/mailman/listinfo/gluster-users
>>>>
>>>>
>>>
> Thank you for  your request.
> But I'm already set the option: network.ping-timeout as the min value,is
> there another options work on the problem ？
> thank you for your help!
>