High await (iostat) on not loaded system

[ciacho]

Hello,

Often this message appears at random agents:

Code: Select all

Your session has been paused


mtop says that's no slow queries but iostat, await column is strange:

Code: Select all

#iostat -kx /dev/sda
Linux 2.6.34.10-0.4-default   03/23/12        _x86_64_        (4 CPU)

avg-cpu:  %user   %nice %system %iowait  %steal   %idle
           5.48    0.76    2.09    2.61    0.00   89.07

Device:         rrqm/s   wrqm/s     r/s     w/s    rkB/s    wkB/s avgrq-sz avgqu-sz   await  svctm  %util
sda               2.67    34.26    0.93    9.17    26.95   174.28    39.86     2.95  291.97   8.08   8.15


await is always close to 300 and it's no matter whether the system is loaded with or without agents.

Code: Select all

Server Performance Report                            2012-03-23 17:18:16
Time range: 2012-03-23 08:00:00 to 2012-03-23 17:00:00

---------- TOTALS, PEAKS and AVERAGES
Total Calls in/out on this server:              9039
Total Off-Hook time on this server (min):    4606.63
Average/Peak channels in use for server:     62.6016 / 107
Average/Peak load for server:                24.6410 / 438
Average USER process cpu percentage:          7.4225 %
Average SYSTEM process cpu percentage:        2.1978 %
Average IDLE process cpu percentage:         90.4071 %


ViciBox Redux v.3.1.13 64 bit
VERSION: 2.4-349a
BUILD: 111122-1333

AGC:
VERSION: 2.4-338c
BUILD: 111114-0039

Hardware:
HP ProLiant ML110 G6
1x Intel(R) Xeon(R) CPU X3430 @ 2.40GHz


I think there may be a problem with hard drive. What do you think?

[ciacho]

Can you check what values ​​await You have?
Thanks!

[fibres]

Hi

I have the following


Linux 2.6.34.7-0.7-pae (dialerm4l) 03/31/12 _i686_ (8 CPU)

avg-cpu: %user %nice %system %iowait %steal %idle
3.34 0.03 0.87 5.58 0.00 90.18

Device: rrqm/s wrqm/s r/s w/s rkB/s wkB/s avgrq-sz avgqu-sz await svctm %util
sda 0.12 57.13 8.06 39.51 213.47 387.07 25.25 5.10 106.63 5.24 24.94

What makes you think it is a database or io issue causing this?

Agents being paused is usually caused by network issues. It happens usually when the web client does not receive the keepalive messages from the server. The server assumes that the client has got disconneted and pauses the agent.

What is your network setup? Is the vicidial in the same subnet?

Regards