Dear Smart People,

 

I've been running unRAID for 5 years or so and have a server pro license.  I upgraded my hardware several months ago and all has been fine.  It may be coincidence, but since I upgraded to 6.3.5, I've been having intermittent server lock ups.  Meaning the UI isn't accessible and my Windows 10 Pro VM isn't accessible and I cannot telnet into the server.  The most recent occurrence was last night.  I typed "diagnostics" from the command prompt on the console while the UI was locked up and have attached the file here.

 

I'm using the Windows 10 VM primarily for running Blue Iris for recording IP cams, and when the server locks up, I don't have any ability to record video.  The server typically runs about 25% CPU usage due to the VM and Blue Iris.

 

I've had an (I think) unrelated issue of my router locking up recently.  That couldn't be related, could it?

 

Can anybody tell me what direction I should start looking to prevent the server from locking up?

 

Thanks in advance for your time and help.

diagnostics-20170927-1715.zip

First suggestion would be to convert all reiserfs disks to xfs, it's a know cause of the problem you describe for some users with latest unRAID releases, and even if it doesn't help you should convert anyway as reiserfs is dead and has terrible performance in some situations.

Thanks for the reply!  I'll read up on the conversion process and have it done by the weekend, at the latest.  If you think that could be the cause of the server freezing, I'll do that ASAP and report back with a new diagnostic.zip if the issue repeats. 

Wow, so I just was reading https://forums.lime-technology.com/topic/35815-re-format-xfs-on-replacement-drive-convert-from-rfs-to-xfs-discussion-only/ and see it's quite the process to convert from reiserfs to xfs.  It'll take me a couple of weeks probably.  I need to see if I have a spare drive to even do this.  

 

What will happen if the server freezes in the middle of a disk copy?  I guess I'll have to cross that bridge when I come to it. 

 

1 minute ago, gatorguy said:

What will happen if the server freezes in the middle of a disk copy?

 

As soon as you have one disk xfs direct all new array writes to it, if it's the reiser issue it only manifests itself when there are writes to a reiser disk.

So to be clear, the destination drive doesn't necessarily have to be bigger than the source drive as long as there's more free space on the destination drive than data on the stored drive.  Is that correct?  For example, I have one 8TB WD Purple drive in my array that I use exclusively for writing security camera footage to.  It currently only has 340GB of data.  My other data drives in the array are 3TB and 4TB drives and I have an empty 4TB drive.  I wouldn't need to buy another 8TB drive to do all the converting, right?  In other words, I don't HAVE to copy from 8TB to another 8TB even though there's only 340GB of data on that drive, right?  I can just start copying data amounts from large to small as long as the empty, xfs formatted destination disk has more free space than data on the source disk.

 

Hopefully I made sense with those words.

3 minutes ago, gatorguy said:

So to be clear, the destination drive doesn't necessarily have to be bigger than the source drive as long as there's more free space on the destination drive than data on the stored drive.  Is that correct?

 

Correct.

Great, thanks!  I'll start converting tonight then.  No hard drive purchasing required!

I'm mostly done with converting the data drives to xfs.  However, last night the server locked up again.  Also, my router locked up again.  Can you tell by the diagnostics zip if the router locking up is causing the server to lock up?  In the syslog.txt which is part of the New diagnostics.zip from this morning (attached below), there are a lot of Ethernet adapter reset unexpectedly.  Then around 6:50am, there's something to do with Ethernet saying "detected hardware unit hang".  I'm not sure what all that means, but that's why I'm asking if somehow when my router locks up if it's causing the server to lock up also.  Any thoughts?

diagnostics-20171002-0744.zip

On 9/28/2017 at 9:20 PM, johnnie.black said:

 

As soon as you have one disk xfs direct all new array writes to it, if it's the reiser issue it only manifests itself when there are writes to a reiser disk.

 

I'm not new to unraid but how do we direct all writes to a particular drive?

 

20 minutes ago, fysmd said:

 

I'm not new to unraid but how do we direct all writes to a particular drive?

 

 

Going to your share(s) and including only one disk, existing data will still be available for access but any new writes will go to that disk.

I tried this and lost the content from the rest of my disks - about to retry

 

to avoid more RFS issues I tried to exclude all RFS disks from all my shares - is this the same thing?