[sdnog] issues with high IO systems

[Samir Abdullatif]

Hi Patrick/Daniel


sorry for the late reply


0.6 TB of data are written each day and are backed up on the shared folder.
the shared folder is part of a backup solution which I don't have admin access for (it's managed by different team who won't change anything from their side)


during certain periods the machine becomes unresponsive even if I perform snmpwalk locally I get timeout.


Regards,
Samir


________________________________
From: Patrick Okui <pokui at psg.com>
Sent: Saturday, December 2, 2017 2:57 AM
To: Samir Abdullatif
Cc: sdnog at sdnog.sd
Subject: Re: [sdnog] issues with high IO systems

On 28 Nov 2017, at 13:14 EAT, Samir Abdullatif wrote:

> Hi all
>
>
> I have 5 servers which are sharing an NFS monted storage for access
> logs. the log files are huge about 0.6 TB of Daily logs.

0.6 TB every day? Wow! That’s quite some data. Or you mean 0.6 TB over
time?


> my issue is that the shared storage is keeping the machines busy and
> not replying to my cacti snmp polling which keeps sending down alarms.


Is the problem speed of NFS or size of the open files?


> any advice on how to go about this issue ?


If the servers are busy waiting for network, then move from NFS
otherwise, rotate the logs.

--
patrick
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.sdnog.sd/pipermail/sdnog/attachments/20171204/6d4b3e77/attachment.html>