Mongodb service stopped

I get why you’re looking to automate restarting mongod, but it might be worth diving into why it’s stopping in the first place before focusing too much on auto-restarting. Otherwise… you’re treating the symptom and not the cause.

Here’s what I would look at first to try get to the bottom of the probs:

1. MongoDB Logs: The first place to look is the MongoDB logs. Check /var/log/mongodb/mongod.log (or wherever your logs are configured). Look for anything odd right before it crashes—OOM errors, disk issues, permission problems, etc.
2. System Logs: You’ll want to check your system logs too. If you’re on Linux, you can use journalctl -u mongod to see if the system is killing it for some reason, like low memory or a hardware issue.
3. Memory and CPU Usage: Sometimes MongoDB crashes because it’s gobbling up too much memory or CPU. Running htop or top can help you see if it’s getting maxed out. If the system’s running out of memory, the kernel’s OOM Killer might shut it down.
4. Disk Space: Make sure you have enough free disk space. MongoDB doesn’t like running out of space, and that can cause it to stop. Just run df -h to check. Also, check that MongoDB has the right file permissions for the log and data directories.
5. Out of Memory (OOM): If your system is low on memory, MongoDB could be getting killed by the OOM killer. You can check this by running dmesg | grep -i 'killed process'. If that’s the case, you might need to increase your swap space or adjust MongoDB’s memory usage in the config file.
6. Check for File Descriptor Limits: MongoDB can open a ton of files, and if your system’s file descriptor limit is too low, it could crash. You can check that with ulimit -n and bump it up if needed.

so… while setting up auto-restart is a quick fix (which you can do with systemd), diving into why it’s happening in the first place will likely alleviate the need to do unnatural things to keep it running.

Let us know if you find anything “telling” in the logs!