Gaps in SVC Graphs

Zoltan

Hi,

recently we are having gaps in one of the SVC graps:



What could be the problem?

Regards,
Zoltan

Pavel

Hi,

Do you use latest 2.20 version? if not the upgrade at first.

How many nodes it has? 8? if so then 10 minute sample interval switch will help.
Just confirm it, I would let you know how to do it.

Zoltan

Hi,

yes, we are using the latest version. We have 6 Nodes in the cluster. If you think we still need to switch the interval, please let me know how to do it.

Pavel

Hi,

Edit etc/storage-list.cfg (10 minutes parameter at the end)
svc-alias:SWIZ:192.168.1.1:
-->
svc-alias:SWIZ:192.168.1.1::::10

kill svcperf for svc-alias
ps -ef| grep svcperf| grep svc-alias

Let it work and let us know.

Zoltan

Hi Pavel,

I have applied the parameter, but I cannot see any change, we still have the gaps in the graph.

Pavel

assure it really work for 10 minutes.
open tmp/<svc-alias>*perf*txt

There is timeframe in the header of each section like:
        Interval Length:  300 seconds

is there 600 seconds?

Zoltan

It looks like this:

CPU-Node Level Statistics
        Interval Start:   2018-02-28 10:37:31
        Interval End:     2018-02-28 10:42:31
        Interval Length:  300 seconds
---------------------
Node,Time,Interval,CPU Busy,CPU Limited,CPU Utilization - System,CPU Utilization - Compression,
STANTSVC11,2018-02-28 10:42:31,300,27.738,0.000,27.738,12.708,
STANTSVC13,2018-02-28 10:42:31,300,38.793,0.000,38.793,22.502,
STANTSVC15,2018-02-28 10:42:31,300,34.170,0.000,34.170,17.965,
STARSSVC12,2018-02-28 10:42:31,300,40.231,0.000,40.231,23.902,
STARSSVC14,2018-02-28 10:42:31,300,43.668,0.000,43.668,16.847,
STARSSVC16,2018-02-28 10:42:31,300,30.280,0.000,30.280,13.333,

Pavel

it still uses 5 minutes.

grep <svc-alias> etc/storage-list.cfg
ps -ef| grep <svc-alias>
date

Zoltan

Here is the output:

grep stsvc10 etc/storage-list.cfg
stsvc10:SWIZ:10.5.118.193:/home/zenoss/.ssh/id_dsa:::10

ps -ef | grep stsvc10
zenoss   11291     1 60 00:00 ?        07:52:51 /usr/bin/perl -w /opt/serviced/volumes/stor2rrd/bin/svcperf.pl stsvc10
zenoss   32404 26292  0 12:59 pts/0    00:00:00 grep --color=auto stsvc10

date
Wed Feb 28 12:59:16 CET 2018

Pavel

1. have you set it in storage-list.cfg today or yesterday?
2. product version.

Zoltan

1. yesterday
2. Version 7.8.1.4 (build 135.5.1712071656000)

Pavel

weird, it should force it to 10 minutes, just tested it and it worked.
try to kill it and check it out after 15 - 20 minutes.

kill 11291    
vi tmp/stsvc10-perf-last.txt

there should be:
        Interval Length:  600 seconds

Zoltan

Unfortunately no change:
 Interval Start:   2018-03-02 07:58:40
        Interval End:     2018-03-02 08:03:40
        Interval Length:  300 seconds

Zoltan

Now it is a bit better, but we still have a gap between 22:50 and 23:30 every day.

Pavel

send us

tail logs/error.log-<storage name>

tail logs/output.log-<storage name>

https://upload.stor2rrd.com or email to our support