Percona restore process stuck in middle

MongoDB Percona Backup for MongoDB
1

Hi, I am using below versions of percona & mongodb in GCP Redhat VM

  1. pbm-agent version - 2.6.0

  2. pbm version - pbm cli - 2.6.0

  3. mongod --version - db version v7.0.11-6

  4. mongosh --version - 2.2.10

I did a full physical backup of MongoDB which was around 140 GB. Then i created a new similar MongoDB with 2 replicas (1 primary & 1 secondary), and tried to restore the backup.

pbm config --force-resync

Then tried restore from backups that I created previously from other MongoDB instance. It started the restore process, went about copying 29GB of data & then got stuck post that.
I tried multiple times with different scenarios (like taking another fresh backup & then restore), but I’m getting into the same state where the journalctl -fu pbm-agent logs are stuck.

1 thing to note that after these 29gb of restore that happens, all of a sudden the CPU spike goes down to almost 0.

There are no error logs or disk related issues nor my VM has any CPU or memory related spike.

I have no clue on what to do, how to proceed further with my restore process.

Can I please get some help on this to find out the RCA?

2

Hi, first of all I suggest testing with a newer PBM version. There have been many bugs fixed since 2.6 and extra logging was introduced which might help pinpoint the problem.

3

Hi, thanks! It did work. I took a fresh backup with upgraded 2.12 version & restored with same version & it worked.

4

glad to hear it helped