[SOLVED] Array not starting (5.0.5)

[andyjayh]

My server has been running quietly in the background for years but I want/need to upgrade to v6. However I have had a faulty parity drive for a while now so thought I better sort that first. The parity drive shows numerous errors and is red balled but if I restart the server it will say that a new party drive is found via the menu, I select the same drive and then I'm off. So I did this and it seemed fine but never really came back to life. The next day I checked to see if there was any activity on the server and as there wasn't I restarted it. Now it starts and mounts the flash drive but doesn't mount the array. Also the web page for the main menu no longer starts so I can't connect to 192.168.1.5:Main

 

I'm scratching my head a little as it's a long time since I have played with this so I'm hoping for a little fault finding guidance. 

 

I'm on the sever terminal and thought I would check the 'go' file to stop and packages starting as each time I restart it is trying to download the latest version of these. However, the 'Go' file only has a line to start the Management Utility, /usr/local/sbin/emhttp &, the rest is commented out. Is this right?

 

The go.bak file looks like this;

/usr/local/sbin/emhttp &

/boot/unmenu/uu

cd /boot/packages && find . -name '*.auto_install' -type f -print | sort | xargs -n1 sh -c

 

If I manually type /usr/local/sbin/emhttp & I get the following;

[1] 25587

Then if i type /boot/unmenu/uu I get the following;

[1]+ Segmentation fault (core dumped) /usr/local/sbin/emhttp

 

I'm probably doing the wrong thing here but I was trying to bring up the main management web page

 

Any pointers for what I should be doing to start the server with no packages and check the array so I can try and start or repair?

Edited January 20, 2018 by andyjayh
Solved

[trurl]

The go file that only loads emhttp is the correct one. It will not load any packages. It's possible you have plugins that are loading things though. Those would be in the plugins folder or the config/plugins folder. If you rename those folders no plugins will be loaded.

 

[trurl]

The most important thing is knowing which disk is parity. If you know that it is pretty easy to start over on V6 with your data intact.

 

You might also post your syslog and SMART report for any disks you want help with.

[andyjayh]

Right, thanks. I remember now that was the change from previous version. I will rename the Plugins directory and try again.

 

[andyjayh]

I have a note from previous which tells me the parity disk. I have also bought a new disk to replace this with, which was the job I was going to do first.

[andyjayh]

Server restarted and menu back. Currently doing a parity sync so I will leave well alone for a while. Thanks for the help so far.

 

Once this is up, should I set about replacing the parity disk before the v6 upgrade?

[trurl]

2 hours ago, andyjayh said:

Once this is up, should I set about replacing the parity disk before the v6 upgrade?

Can't advise how critical any disk replacements are without SMART reports.

[andyjayh]

Fair enough. Will let the current party sync finish then look to extract a SMART report.

[andyjayh]

Just for completeness, party check finally completed yesterday. V6 upgrade now done, much easier than I was expecting

 

Party disk is already showing up some SMART errors so that will be my next job. Trying to read up and understand the correct process. Should i just shut the array and server down and replace the parity drive with the new drive and let unRAID rebuild onto the new? Or do I need to introduce the new drive in its own location and format & pre clear before then placing into the party drive slot?

[andyjayh]

This is the process I'm about to follow;

 

Upgrading parity disk(s)

If you wish to upgrade your parity device(s) to a larger one(s) so you can start using larger sized disks in the array, the procedure is as follows:

1. Stop the array.
2. Power down the unit.
3. Install new larger disks.
4. Power up the unit.
5. Assign a larger disk to the parity slot (replacing the former parity device).
6. Start the array.

[bonienl]

In V6 It is as simple as shutting down your system, replace the old parity disk with a new disk. Start your system and select the new disk as parity. This will start automatically a parity rebuild once you start the array.

 

Btw the same procedure can be used for any data disk (assuming the new data disk is not larger than the current parity disk)

Edited January 20, 2018 by bonienl

[trurl]

Even though you have already found the procedure, and received a reply, I will go ahead and post this for further edification since I was already writing it.

 

Now that you're on V6 go to Tools - Diagnostics and post complete zip. That will let us look at SMART for all drives. Unless you are very confident in the SMART of all other drives it might be a good idea to let us see them.

 

Be sure to setup Notifications if you haven't already so unRAID can notify of any future problems.

 

unRAID only requires a clear drive when you are adding it to a new data slot in an array with valid parity. This is so parity will remain valid since a clear data drive has no impact on parity. If you don't preclear a drive, unRAID will clear it for you. Unlike previous versions, V6 doesn't take the array offline when clearing.

 

So, your scenario doesn't require a clear disk. And you would never format a parity disk anyway, since parity doesn't have a file system. People think format is some generic "prepare this disk for use". Format actually means "write an empty filesystem to this disk".

 

Some people prefer to preclear a disk just for testing even when a clear disk isn't required, but not everyone agrees there is much benefit to this. In your case, it might be better to get the parity disk replaced quickly, though I can't say for sure since I haven't seen SMART for it or any other disk.

[andyjayh]

Thanks, really appreciate your help and advice on this.

 

Run the diagnostic and attached the .zip to this post.

media-diagnostics-20180120-1140.zip

[JorgeB]

Parity disk is failing badly, other disks are mostly OK but at least 2 or 3 might need improved cooling, unless these are very old values:

 

Lifetime    Min/Max Temperature:     20/70 Celsius

Lifetime    Min/Max Temperature:      0/67 Celsius

Lifetime    Min/Max Temperature:     20/55 Celsius

 

You'll want to keep them under 40C for best lifespan , 45C tops.

 

 

[andyjayh]

Thanks for that feedback. The drives are in cooled drive bays but i suspect this is not enough during the summer months as my server sits in my loft and it gets hot up there. Will have a rethink on location of the server.

[andyjayh]

Although, I've just noticed that the fan on one of my driveways has failed! Funny enough, the one that contains the party disk. I will look to see if it is also the other drives highlighted above. New drive caddy required!

[andyjayh]

Fan now swapped with the unused drive caddy so all drives are being cooled. Thanks for noticing this