Disk Missing Advice

[tw332]

Hi,

 

I've suffered issues with my server not responding on the web interface after a while of being online (sleeping daily and woken when needed). I can usually then log in via SSH and attempt a safe power down but that usually also freezes. (But that's a topic for another day.)  At this point, unfortunately, it's a hard reset - which I had to do on Tuesday night.

 

Server rebooted and at some point on Wednesday night, I noticed my shares were missing (presume the automatic parity check failed). Logging into the interface I noticed one of my disks was reported as missing... also noticed my email notifications weren't working! As I've not done any config with unRaid for a long time I completely forgot to save the log before rebooting.

 

Rebooted - same issue. Did a quick bit of trouble shooting in the interface and in SSH and realised my mistake with the log. As I didn't have time to check cables and connections and as the drive was old and only placed in the server as a temporary measure (18 months ago, with the intention of replacing soon after)... I decided to order a new disk (which I've received today) and shut the server down to prevent any further risk.

 

When I got home from work the next day, my server was woken automatically by magic packet. The server booted - disk was found and a parity check had been started.  I cancelled the parity check and started a new one with the option to not write changes to the parity disk (I  was unsure on the failed disk and wanted to see what unRaid would find without changing anything in parity).

 

Zero errors were found - which is good.

 

My question is (and apologies for the long winded post to get to it) - what's the best for me to do now?

 

1) pre-clear the new disk once and replace the older, possibly failing disk as soon as possible.

2) pre-clear the new disk several times with a view to replacing the older disk within the next month (and having it available in an emergency if the older disk does fail).

3) pre-clear the new disk and leave the older disk running until it fails.

 

It's been quite a few years since I set up the server so apologies if any of this is outdated... (I am planning to upgrade to the newer versions sometime in the New Year).

 

Thanks,

Matt.

[trurl]

Since you have posted in the v5 area, see the sticky at the top of this subforum that tells you how to get syslog and smart reports.

[tw332]

For a moment I thought you were pointing me in the direction of obtaining historical logs!

 

For anyone that's interested, I've performed SMART tests etc on the missing disk and have not been able to identify any issues. Server is operating as it always did. I've precleared the new disk and will be replacing the (non)missing disk with it soon.

 

Would there be any preference to upgrading to the latest version before rebuilding on to the new disk or after. If it matters, I'm still on a v5 RC.

 

Otherwise, the topic can be closed.

[trurl]

Best to get your array stable before upgrading. First rule of troubleshooting, only change one thing at a time.

[tw332]

Thanks - I'll get started on that.