Morning, 

First off, I ignored the instructions in the documentation regarding preclearing drives before installing. "What's the chances this will affect me?" Hubris came to me this morning. 

Situation: I have replaced my two Parity drives at the same time, setup a 'new config' while maintaining array devices, the two old parity drives were then removed and added as array drives, replacing 2 x 1TB drives that did have data on them. This morning the Parity sync failed/paused due to write error's on one of the drives. 

I resumed the sync with the error'ed drive offline so that I have one drive up and running for coverage before looking into this disk to see if I have a bad egg or something else. After resuming I noticed that docker wasn't running, and was failing to start. Going to the logs, I could see that NGNIX was reporting out of shared memory errors. Realised that when I removed one of the drives, it had my docker images and data on it. Don't know how or why ngnix was starting without an image or data but \o/. 

The interesting thing was 2-3 minutes after these memory error's stopped, is when the disk errors started. It seems like a weird coincidence of interchange of errors? 

 

I'm waiting for the Sync to finish before doing much more troubleshooting, according to the very accurate estimated time remaining its just under 2 hours. Mainly wanted to post up incase the shared memory flagged an issue or further troubleshooting required.
 

carvahall-syslog-20250308-2318.zip

Attach Diagnostics to your NEXT post in this thread. 

Attached. 

carvahall-diagnostics-20250309-1143.zip

Parity dropped offline, looks more like a power/connection issue, but since there's no SMART, check/replace cables and post new diags.

Sorry for the delay, was restoring some backups and stuff. 

Since last post I have done the following: 

Stopped array. Couldn't the normal drill down settings screen where you start the smart test for troubled drive. 

Started array. AB

Stopped Array. AB

Started in Maintenace. AB

Rebooted server. 

Performed smart test (short).

Said that it finished without error. There were 100 reallocated sectors and another that I can't remember right now. 

Re-assigned the drive to P1 and I have started the rebuild. 

 

carvahall-diagnostics-20250309-2222.zip

2 minutes ago, MaxITBarcaldine said:

Performed smart test (short).

Many drives pass the short test that would fail the extended test so it should only be treated as definitive if it fails.

 

3 minutes ago, MaxITBarcaldine said:

There were 100 reallocated sectors and another that I can't remember right now

This is never a good sign.   Reallocated sectors are not an issue as long as the number remains stable, but you have enough that they may be early signs of incipient problems with the drive

58 minutes ago, itimpi said:

Many drives pass the short test that would fail the extended test

Yeah, parity is showing pending sectors, you should run an extended SMART test.

Final update by the looks. 

Parity rebuild failed after 15 hours. Now the disk isn't showing after reboot. 

The drive itself is plugged into... a thingy you get with servers. Where is splits off into multiple sata/power connecters to plug in drives. One of the connecters was already broken so I don't have a spare to swap to check. 

Sata ports on the MB don't seem to work either? Maybe a configuration issue in bios but I couldn't be bothered troubleshooting on the main server. I'll pull the drive next weekend and load it onto a test bench to get some SMART results. 

Last thing I'm hoping for help with is help finding the right part for replacement. 
So i have HP Proliant ML150 Gen 9, I have one set of drives that have caddies that I can slide in and out from the front. The other is the default cage and is a PIA to remove. Would like it replaced with similar, but google is only finding me drive trays. I realise this is pebkac problem and I'm not searching the right thing but I'm not sure what I should be searching.