Offline Reorg?

[Ryan Zerbe]

Hello,

A few years back we had a DU node audit from CA/Automic, dont remember the timeframe. We sent our trace files in, CA/Automic checked for best practices, etc. One of the findings was to run a quarterly-ish offline reorg. I have a job developed to do this work, some of my nodes have it running without issue, others not. The issue is finding the best time for each node to be down for a few mins so nothing is interrupted. We originally decided to institute this because a few nodes would hang up and not execute anything until we shutdown DU node, ran an offline re-org and restarted. We have since not had that problem anymore, but we also dont have the problem on any of the nodes where this script is not running.

TL;DR, is it still recommended/advised to run a monthly/quarterly offline reorg? I believe we were on DU v6.4/6.5 or so at this time, we are now on v6.7 and v6.10

Thanks for any info.

[Eric Delawoevre]

Hello,

here is my point of view

the V6 reorganization is much more efficient than in v5.
It takes a few seconds in v6, while several 10 minutes in v5 sometimes.
But it is true that stopping a node is always very problematic for our loving production plans :-)

Consider that:
* Machine restarts are almost mandatory for security reasons (security patch). Deviation from running machine systems (more frequent in Windows than on Unix).
* an offline reog is nothing more than shrink data in flat files.
* update your node regularly (release each semester approx)

Thus, stressing your production plan with an offline reorg is too much.

It is important to let your v6 node hot reorg ,during low activity (node settings)
I know that all my nodes made a cold reorg but never only for this goal

regards
Eric

------------------------------
Main director
SynApps
#dollaruniverse
------------------------------