Pool Tbl is empty

dettmari

We just implemented STOR2RRD for our V3700 storage. It was working fine until we adjusted the tiers of the pools. Now on the report heatmap I can see only some LUNs on Pool2 and nothing about Pool1 anymore. Plus the Pool Tbl is empty. Everything else works fine, no problems. What could cause this?

 

Lukas

Hello,

There was a small bug. Here is the fix:

https://download.stor2rrd.com/patch/7.13-1-gb64d/heatmap-stor.pl.gz

Gunzip it and copy to /home/stor2rrd/stor2rrd/bin (755, stor2rrd owner)

-rwxr-xr-x 1 stor2rrd stor2rrd 168581 Jun  9 09:51 heatmap-stor.pl

If your web browser gunzips it automatically then just rename it: mv heatmap-stor.pl.gz heatmap-stor.pl

Assure that file size is the same as on above example

after copy script to bin directory

cd /home/stor2rrd/stor2rrd # or where is your STOR2RRD working dir

. etc/stor2rrd.cfg

$PERL bin/heatmap-stor.pl

refresh GUI

If the fix did not help you send us logs.

cd /home/stor2rrd/stor2rrd # or where is your STOR2RRD working dir

tar cvhf logs.tar logs tmp/*txt

gzip -9 logs.tar

Send us logs.tar.gz via https://upload.stor2rrd.com

Thank you

dettmari

I get some errors when running the Perl script: 

bin/heatmap-stor.pl: line 2: use: command not found
bin/heatmap-stor.pl: line 3: use: command not found
bin/heatmap-stor.pl: line 4: use: command not found
bin/heatmap-stor.pl: line 5: use: command not found
bin/heatmap-stor.pl: line 6: use: command not found
bin/heatmap-stor.pl: line 7: syntax error near unexpected token '('
bin/heatmap-stor.pl: line 7: 'use JSON qw(decode_json);'

After running it, I can now see Pool2 in the Pool Tbl but Pool1 is still gone.

Lukas

Hello,

cd /home/stor2rrd/stor2rrd # or where is your STOR2RRD working dir

./load.sh heatmap

refresh GUI

Thank you

dettmari

Hi, I couldn't get the script to run without error, so we decided to reinstall the system and the heatmap works fine now. However, after the reset, I have a problem that the data is one hour behind. We had a wrong time zone on the storage (GMT+8 instead of GMT+7), which we already corrected. But the data is still lagging one hour behind. How to fix that? Shall I open a new discussion for it?

Lukas

Hello,

you don't have to. What version of stor2rrd do you have? please these outputs:

ssh stor2rrd@[svc|storwize control enclosure IP address] "svcinfo lssystem" | grep time

timedatectl

Thank you

dettmari

Please see here

dettmari

And this is the problem, all graphs are 1 hour behind.

Pavel

Hi,

cd /home/stor2rrd/stor2rrd

date

ls -l tmp/<storage name>*txt

head tmp/<storage name>-perf-last.txt

dettmari

dettmari

Hi, any advice about this?

dettmari

We still didn't find any way to get rid of the 1 hour time difference.
Any hint is appreciated. 

Pavel

Hi,

1. screenshot of the actual time from storwize UI

2. actual time on stor2rrd server

dettmari

I hope that shows the time is the same.

Pavel

Hi,

1. is that hapen for 1 storage only or for more

2. send us log, note storage alias name and local time when logs were captured

3. attach this output

cd /home/stor2rrd/stor2rrd

ls -l data/<storage alias>

Note a short problem description in the text field of the upload form.

cd /home/stor2rrd/stor2rrd # or where is your STOR2RRD working dir

tar cvhf logs.tar logs tmp/*txt

gzip -9 logs.tar

Send us logs.tar.gz via https://upload.stor2rrd.com

dettmari

I have uploaded the logs as requested. The problem only happens for one storage (V3700). We're monitoring two and the other (DS3500) works fine.

Pavel

Hi,

time on the storage and stor2rrd time is shifted about 1 hour for sure.

SVC CLI svqueryclock: Wed Jun 30 17:35:57 NOVT 2021

stor2rrd server: 2021.06.30 18:35:57 WIB +0700 1625052957

TZ on the storage  is fine: "Asia/Novosibirsk" - UTC +7, stor2rrd server has same one: UTC +7, both using NTP

In any case we cannot do anything, raise a call with IBM support that svqueryclock is not right although TZ is fine.

Let us know.