Can you provide the deployment logs?

if you can, share the deploy logs because "mongo crashed" can be a few different things.

but first thing I would check is whether the log says something like MongoDB cannot start on Linux kernel 6.19+. if yes, redeploying wont fix it, it is the Mongo 8/kernel issue. in that case I would try a Mongo 7 image, or move it to a region where the host kernel does not hit that issue.

if the log is WiredTiger / data files / lock file, dont delete the volume. that is a different recovery path.