Error current heartbeatlag (9223372036.85s) is too high

[ritxman]

Hi all, I want to ask about gh-ost migration.
When I run gh-ost migration on my virtual machine and the progress reached 100%, it got error log current heartbeatlag 9223372036.85s is too high, it needs to be less than both --max-lag-millis (1.50s) and --cut-over-lock-timeout-seconds (3.00) to continue.
Here is the current specification:

1. gh-ost version is 1.1.6
2. using MariaDB version 11.4

What I have done are:

1. Upgrading the gh-ost version from 1.1.5 into 1.1.6
2. Running the gh-ost command without --max-lag-millis
3. Reseting the master database to refresh the replica
4. Run ANALYZE TABLE to refresh the performance schema
5. Also I have set the timezone setting into UTC+7 to comply with my current timezone in Indonesia by running this SQL: SET GLOBAL time_zone = '+07:00'; and SET time_zone = '+07:00';

So I'm curious about the printed value of TimeSinceLastHeartbeatOnChangelog(), I decide to debug it on my local by print its value, but it shows 0001-01-01 00:00:00 +0000 UTC, hence the heartbeat lag value become extremely huge.

Do you have any idea why this happens? Will appreciate any help given for this issue.
Thanks in advance!

Attachments:

[omkarkulkarni2704]

Ensure performance_schema is enabled and set --heartbeat-interval 1s in your gh-ost command to fix the heartbeat lag issue.

• Ensure performance_schema is configured correctly:

SET GLOBAL performance_schema = ON;

• Add the --heartbeat-interval option in your gh-ost command:

--heartbeat-interval 1s