Unexplained replication lag #2431
Replies: 1 comment
-
I did some research I noticed there are the following issues opened 2003 and 1814. I understand that replication lag can happen on idle databases since it is a difference between the last visible commit to the current time. The LSNs are the same so the replicas are aligned with the primary instance. However, I don't understand why that issue is not always occurring, on another test database which is also idle database we don't experience that behavior. It is important for me to understand that metric since we are moving the production workload to the operator. I would appreciate it if you could shed some light on it how to interpret that metric and related issues with that. Thank you, |
Beta Was this translation helpful? Give feedback.
-
We are running a few databases for testing purposes. On all of them, we are getting an alert concerning the replication lag that is constantly increasing. As I mentioned those databases are created for testing purposes and the change on them are happening occasionally. The following query executed on one of the replicas gives the following result:
Any DDL query running on one of those databases is zeroing that lag. I have just created a database, the LSN’s have been changed and the lag has been started to grow from zero.
The CNPG status does not show any additional information. The streaming reaplication is working correctly, no lag has been discovered.
As you can see the replication lag is constantly growing,
Could you please help me to understand what is the root cause of that replication lag and how we can solve it?
Thank you,
Beta Was this translation helpful? Give feedback.
All reactions