We had the same issue, in our case the issues was caused by some corruption and the sql daemon process locking the latch. Also other issues we seen in this system was that we were not able to take a full backup or diff backups. Even though we could take tlog backups the pages were not getting release and tlog was getting large.

The steps we follow are:

Verify the backups and can restore the db using tlog backups

Verify the db is up-to-date

If there is a problem with the backups (in our case)

Stop all the connection to db

Restart the db

This will put the db in to recovery mode (this may take some time ….)

Backup the database

Check db, dbcc checkdb

if you still getting the same error, timeout on buffer……

restore the db using the backup

check the restored db

fix and errors

backup the db

put the db back to production

hope this help