SYNCIQ - Isilon - we cannot stop sync IQ job is a way to force to stop

Hi Arek,

Did you stop the synciq services on both the source and target clusters?  The workers should eventually timeout...

Cheers,

Matt

There is a issue with syncIQ version 7.0.2.9 that we are running on Isilon. Right now Isilon support is working on a patch to fix Isilon reporting database issue but they are not sure when patch is going to be available for release.

In mean time if anyone sees any issue where web interface is running slow please log on via CLI to Isilon and try running below mentioned commands

NYHQ-NAS-ISL02-1# find /ifs/.ifsvar/modules/tsm/sched/reports -name "report-[0-9]*.gc" | wc -l   ( This command will show you size of the reports)

find /ifs/.ifsvar/modules/tsm/sched/reports/ -mtime +30d -name "report-[0-9]*.gc" -exec rm -vf {} \; ( This command will remove reports that are older than 30 days, you can modify the days if there are lots of report ie 15d for 15 days)

Once thats done and even after that if you see web interface is running slow try running below command.

Run a test syncIQ job and then try to run a report

isi sync policy report -v SyncIQ_Test_NY_to_NJ  (If this command doesn't return anything or take long time to return)

mv /ifs/.ifsvar/modules/tsm/sched/reports/reports.db /ifs/data/Isilon_Support (Run this command to move the current report database to Isilon support folder under /ifs)

python -c 'import isi.fs.siq;isi.fs.siq.SyncIQUtils().rebuild_reportdb();' (Run this command to recreate the database)

Once this is done, web interface should start working fine and any jobs that are running should return normal behavior.

Hi Arek,

I have seen this at another customer as well; it comes from many policies executing often, and reports being retained for a significant amount of time.  The reports themselves have default settings for a 365 day retention and 2000 reports per policy...

In general, if you are running a lot of policies multiple times a day, you might want to reduce those settings to something more reasonable (maybe a 30 day retention and 100 reports per policy?).

Cheers,

Matt