We are running a cluster of 3 couchbase servers in kubernetes on three separate VMs (1 VM per couchbase node). At 00:10 AM we take a full backup and move it out to cloud storage. Then every hour after that we do a differential backup except at midday when we do a cumulative backup.
We have noticed that the backup takes longer and longer each time we run it and that it uses a lot of RAM while it is running. We have to run the cbbackup process on its own node (VM) within the kubernetes cluster.
Our database is only around 1.6GB with about 1.6m documents.
We run cbbackup with the following options:
cbbackup -m <backup type> <host> <backup dir> -u <user> -p <password> -d <bucket name>
We tried adding
-x batch_max_bytes=4000000 -x cbb_max_mb=50000 -x batch_max_size=1000
But these options did not seem to make any difference to the amount of RAM being used or the amount of time the backup takes
We are not far off the backup taking so long that the next hourly backup will start before the current backup has finished.
Any advice on how we might bring the backup time down and also limit the amount of RAM that cbbackup is using?
Thanks in advance