Troubleshooting
Problem
nzstart command uable to start, and you are getting error like the followings: Lock file "/nz/data.1.0/postmaster.pid" already exists.
Symptom
The problem could happened after an software upgrade on a IBM PureData System for Analytics system. Where 'nzstart' command will fail with the following error message:
nzstart: Error: unexpected error accessing system catalog:
Lock file "/nz/data.1.0/postmaster.pid" already exists.
Is another postgres (pid 5907) running in "/nz/data.1.0"
This is not allowing the system to come back up
Cause
postmaster.pid file is a lock file. This file is used to make sure that only a single master process: "postmaster" is running.
In this case, the file was not cleared properly; possibly from the abnormal termination of postmaster process previously; or some other process is preventing the file to be deleted.
When this happened, 'nzstart' fails to startup normally.
Diagnosing The Problem
The system is not online, but when you issues this command:
ls -l /nz/data.1.0/postmaster.pid
If the file existed, you will the get the error message below:
nzstart: Error: unexpected error accessing system catalog:
Lock file "/nz/data.1.0/postmaster.pid" already exists.
Is another postgres (pid 5907) running in "/nz/data.1.0"
Also run 'lsof +d /nz/data.1.0' ; you should not see any process still using that directory.
Resolving The Problem
- Run 'nzstop' to make sure that all processes are stop properly.
- As 'root' user, run 'service nzhealthcheck stop'
- Log out of all nz sessions (CLI and client connections).
- Log back in and check whether /nz/data.1.0/postmaster.pid still exists.
- ls -ltr /nz/data.1.0/postmaster.pid
- Run the following command to check if any process still holding up:
- lsof +d /nz/data.1.0
- Run 'nzstart' again if step 4 & 5 did not returned any result. If not working, please contact IBM Support at http://www.ibm.com/support
Was this topic helpful?
Document Information
Modified date:
17 October 2019
UID
swg21667910