Today in one of the Production Server we were continuously getting an error “Timeout occurred while waiting for latch: class ‘LOG_MANAGER'” and our application almost came to standstill. At early hour, I got an emergency call and had to respond to it quickly.
Looking at the latch wait error it became clear to me that a transaction log is growing because it could not clear\truncate in timely fashion for some reason. Firstly find the database where log is growing and then we can find out what is preventing log clearing by below command:
SELECT [log_reuse_wait_desc] FROM sys.databases WHERE [name] = N’database_Name’
In my case, the log file was inappropriately sized and hence it was increasing frequently which was the main reason for the issue. I sized the log file appropriately and since then we have not faced any issue.
Cheers,
Subhro Saha
Welcome to the Database World... 