Db2

 View Only

 AWS EC2 DB2 HADR: db2cm fails with "Error running command: crm configure property stonith-enabled=false"

Roberto Stradella's profile image
Roberto Stradella posted Fri August 01, 2025 08:04 AM

Hello, I'm in trouble to configure a DB2 HADR Pacemaker cluster in AWS EC2. No way to run successfully db2cm utility. Here a brief description of what happens:

Env: AWS EC2, Linux RHEL 9.4, Db2 version 11.5.9.0 special_58840

Prb: we are configuring Db2 Pacemaker HADR cluster. The db2cm -create -cluster -domain <domain name> -host <host1> -publicEthernet eth0 -host <host2> -publicEthernet eth0 fails with the error:
Line: 1340 Error running command:
crm configure property stonith-enabled=false

all the prerequisite software have been installed on both hosts;

SELinux has been disabled on both hosts and they have been rebooted;

root and instance owner can run ssh passwordless successfully on the local hosts and remote hosts both for long and short hostnames;

no other ideas how to fix the problem.

Has anyone already encountered the same problem as me? Any ideas or suggestions?

Thanks. Ciao. Roberto.

 

Madhusudan S M's profile image
Madhusudan S M posted Fri August 01, 2025 11:51 AM

Hello Roberto,

Run the 'crm' property change failing Command Manually.
Watch for Error or Success Output.
   - If successful, it will return without errors.
   - If it fails, you'll get more detailed output.
Check for any error messages like:
   - crm command not found
   - Unable to connect to the cluster
   - Permission denied.
Additionally, run the Db2 Pacemaker prerequisite checker:

  ./db2prereqPCMK


Thanks
Madhusudan
Damir Wilder's profile image
Damir Wilder IBM Champion posted Mon August 04, 2025 04:18 AM

Hi Roberto,

There's an IBM Support article that looks related to your issue:

https://www.ibm.com/support/pages/db2-pacemaker-stonith-enabled-issues

Checki it out, hope it helps!

Regards, Damir

 

Roberto Stradella's profile image
Roberto Stradella posted Mon August 04, 2025 11:24 AM

Hello, thanks for the help provided. In the meantime I opened a case to IBM (TS019846924) and they provided me the workaround as follows:

"so here if db2cm directly call for ssh command, its works and I found that the upgrade_seq location has changed in crmsh 4.5. Could you test the following and let me know if it changes anything?

Run on both nodes:

mkdir -p /var/lib/crmsh

echo '1.0' > /var/lib/crmsh/upgrade_seq

After running the above commands the db2cm -create -cluster -domain worked successfully.

Ciao. Roberto

Related Content