May 2022
2 / 2
May 2022
May 2022
Tim_Pynegar
May 2022

I’ve got a three node cluster setup running on kubernetes with each node having a ~1TB Drive. I’ve created a Persistent Volume Claim for Mongodb of 100 GB.

When I get the actual oplog size from rs.printReplicationInfo() it’s reporting 45011.683349609375 MB

When I run db.oplog.rs.dataSize() it’s reporting 469617794

Should I pass in the size in the conf file to make it 5% of the size of the Persistent Volume Claim as it seems it’s reading my system drive and in the process taking up 50% of the space in my Persistent Volume Claim?

17 days later
AasawariAasawari SahasrabuddheMongoDB Employee
May 2022

Hi @Tim_Pynegar and welcome to the community forum!!

Do you mind providing more details regarding your setup:

For the above mentioned deployment, is 1 TB shared among all three clusters or is this for all the three nodes.

  • Could you share the deployment/statefulset with persistent volume yaml files for the deployment.
  • MongoDB version for the deployment
  • is the 5% your specific requirement (which perhaps better be set in a config file), or are you discussing about the default MongoDB behaviour?

P.S. here are two things which can be noted for reference:

  1. Oplog can grow beyond their set size when the majority commit point is behind.

  2. The oplog is stored in a database called “local”, and it resides in the dbpath of the instance
    If the dbPath resides in a persistentVolume then the size of the entire database is bound by that volume, including the oplog size.

Thanks
Aasawari


Want to read more? Browse other topics in Replication or view latest topics.