On 8/7/23 08:41, KK CHN wrote:
List ,
Description:
Maintaining a DB Server Postgres and with a lot of read writes to this Server( virtual machine running on ESXi 7 with CentOS 7) .
( I am not sure how to get the read / write counts or required IOPS or any other parameters for you. If you point our I can execute those commands and get the data. )
Peak hours say 19:00 Hrs to 21:00 hrs it hangs ( The application is an Emergency call response system writing many Emergency Response vehicles locations coordinates to the DB every 30 Seconds and every emergency call metadata (username, phone number, location info and address of the caller to the DB for each call)
During these hours the system hangs and the Application ( which shows the location of the vehicles on a GIS map hangs ) and the CAD machines which connects to the system hangs as those machines can't connect to the DB and get data for displaying the caller information to the call taking persons working on them. )
Have you looked at the postgresql log file?
Have you looked in the
application error log?
Issue :
How to trace out what makes this DB hangs and make it slow and how to fix it..
Resource poured on the system :
64 vCPUs allocate ( Out of a host machine comprised of 2 processor slots of 20 cores each with Hyper Threading, intel xeon 2nd Gen, CPU usage show 50 % in vCentre Console), and RAM 64 GB allocated ( buy usage always showing around 33 GB only )
What else happens on the ESXi host between 19:00 and 21:00?
Query :
How to rectify the issues that makes the DB server underperforming and find a permanent fix for this slow down issue.
Attached the Postgres.conf file here for reference .
Any more information required I can share for analysis to fix the issue.
Krishane
--
Born in Arizona, moved to Babylonia.
