Restarting node on cluster fails gvwstate.dat no such file or directory

I had a bad failure due to folder deletion. I logged into my server and tried starting the node. I am getting the following error WSREP: Fail to access the file (/db//gvwstate.dat) error (No such file or directory). It is possible if node is booting for first time or re-booting after a graceful shutdown I checked the directory and the gvwstate.dat file is not there. I do have galera.cache grastate.dat files in the directory.
On the node that is down the grastate.dat file has

# GALERA saved state
version: 2.1
uuid:    00000000-0000-0000-0000-000000000000
seqno:   -1
safe_to_bootstrap: 0

On my donor node the grastate.dat file is

# GALERA saved state
version: 2.1    
uuid:    febfad99-09fe-11ea-ad83-57c73422d738
seqno:   -1
safe_to_bootstrap: 0

Help

1 Like

Don’t worry about the gvwstate.dat file.

On the failed node, simply delete the entire data directory and start mysql. This will force an SST from the donor node. This is the easiest and simplest way to get the node back.

1 Like

In my.cnf I had to use rsync as the --wsrep-sst-method instead of xtrabackup-v2.

1 Like

@danarashad rsync is extremely legacy and should not be used. You should endeavor to correct the issue with xtrabackup.

For some reason using xtrabackup kept failing. So I opted for rsync method.

Yes, I understand that, but rsync is a dead/deprecated method and is only there for legacy purposes. You should post more information from your logs to understand why xtrabackup method is failing.