I have a number of servers running Percona 5.5 and using innobackup to perform full and incremental backups.
On one server the incrementals do not complete. They just abruptly end. When I run the same command a few minutes later outside of the crontab, they run flawlessly.
The incremental command in the crontab is: /innobackupex.log
I am writing the logs to the /tmp/innobackupex.log file and in there I can see only:
log scanned up to (10139137934104)
That is all, the file stops there
As I said, all my other databases are just fine - some do hourly incrementals and it works perfectly.
All databases are different and hold different data, the one with the problem is the largest at over 500GB.
The database being backed up is a slave.
(Note: Incrementals are set up to always use the base and not build on previous incrementals)