My Prod database has taken almost 1 hour to shutdown after the stop command

[Former Member]

Hello Experts,

      My production database of size around 1TB in distributed system with 4+2 nodes. Last week, we tried to bring down the DB for OS patches, the HANA DB took almost 45 mins to come down.

       The BW application server was down few mins before we issue the command in DB side. I am aware of the internal process during shutdown. But my question is any one could able to explain is there is any best practices before issuing the DB stop command in HANA.

     I would like to effeciently handle the DB outage.  Your thoughts will be appreciated.

Regards

Suresh

[Former Member]

Suresh,

I have to seen a problem like this on my installation and I guess the issue comes related with logging and persistance.

how many time it takes to start=?

After all services are started, try to down the server again.

if faster, maybe the logging and persistance are not being backed up properly

if not, the problem is another.

[Former Member]

Pablo,

      Thanks for your response.

       The startup took 12 mins on that day. Yesterday we brought down the prod HANA DB for some parameter change and it was ultra fast. The whole process of restarting was less than 10 mins.

        Not able to visualize why it took so long. Might be due to Delta-Merge mechanisms nd subsequently writing to disk.

Regards

Suresh

[Former Member]

Inside global.ini, which are the values for "log_mode" and "enable_auto_log_backup" ?