I'm having an issue where the server has been rendered unusable.

[FlyingTexan]

Every menu I click on it lags for 30 sec or so then says it can't connect. Another 30 seconds go by then it can. Then it happens again. I simply can't use the system. 

I believe I had my cache drives set incorrectly where I've basically slammed the system with too much. The drives fill then SAB has a ton of things queued or decompressing and the system just locks up.  What log do you need and where can I find it to put here?  I also have gone to tools and disabled docker.  I reboot the system and am running the mover. I now have appdata, downloads, system set to cache prefer then have "Homeserve" the directory for everything I download set to YES.  I rebooted with the docker disabled and have now started MOVER. This is how things are currently sitting.  Even as I type this if I go to my webGUI and click on anything it tells me the site can't be reached. 

 

 

I'm also wondering if it could be hardware related like an issue with the CPU or a hard drive. Is there a way to see the transfer rates of what's going on?  I put system to Prefer cache and it's still showing on disk 6. same with appdata not moving any meanwhile the cache still had almost 500GB free. 

 

How long does mover take? Does unraid write multiple files to multiple drives at once or am I limited to the single drive speed, around 135 MB? What's getting me is even with dockers disabled it's still buggy and timing out at every other click. Where can I get the diagnostics you need? Is there a means to test the different components to see if there's a fault?

[FlyingTexan]

On dashboard I'm not getting any information for the CPU. shows 0% everything. 

[FlyingTexan]

found this option so downloaded

tower-diagnostics-20210622-0809.zip

[FlyingTexan]

[6of6]

The UnRaid WebUI isn't reliable. Can you test if everything is actually working?... the shares, the media, etc.?

 

I ask because my WebUI went "crazy" and then my daughter (after telling her everything wasn't working) said everything is working just fine. She was right; everything was working, except the UnRaid WebUI.

 

6.

 

 

[FlyingTexan]

6 hours ago, 6of6 said:

The UnRaid WebUI isn't reliable. Can you test if everything is actually working?... the shares, the media, etc.?

 

I ask because my WebUI went "crazy" and then my daughter (after telling her everything wasn't working) said everything is working just fine. She was right; everything was working, except the UnRaid WebUI.

 

6.

 

 

It's not just the webUI it's the server's response itself. I've had mover running all night and the system is still not peforming at all like it should. Again the CPU isn't showing any data anymore. 

[FlyingTexan]

Mover did make some progress but it must do one part at a time because it has been freeing up space on the cache by moving "homeserve" off but appdata and system have not been touched at all. 

[itimpi]

13 minutes ago, FlyingTexan said:

but appdata and system have not been touched at all. 

Why would you expect these to be moved?    The ‘prefer’ setting tells mover to move the share contents ONTO the cache if there is any room.     

[FlyingTexan]

If you look at the photos you'll see appdata has 1tb on disks 1-2-3 and all of system is on disk 6.  

[itimpi]

5 minutes ago, FlyingTexan said:

If you look at the photos you'll see appdata has 1tb on disks 1-2-3 and all of system is on disk 6.  

I believe that mover tends to process shares (and their contents) in alphabetical order..   There was probably not much spare space when it started on appdata and do nothing got moved.    You say it is currently moving data off the cache for the ‘homeserver’ share which is freeing up space?
 

The ‘appdata’ share seems to be bigger than the cache can hold which is not a good scenario to be in - any idea why it is that huge?

 

it is also worth mentioning that the ‘system share contents can normally only be moved if the Docker and VM services are stopped as these hold files open and open files cannot be moved.

[trurl]

Go to Settings and disable Docker and VM Manager until you get this resolved. Wait for parity check to complete then post new diagnostics.

 

Which downloading dockers do you use? Are any of them downloading into appdata?

[trurl]

Also, what is your btrfs configuration? I notice you have different sized disks in cache pool.

[FlyingTexan]

5 hours ago, trurl said:

Also, what is your btrfs configuration? I notice you have different sized disks in cache pool.

I disabled docker before ever enacting the mover. I have 2x1TB and 2x4TB.  The cache has been corrected now. I'm still having the disconnect issue from the server. If I click on anything in the server it will lag, disconnect, wait about 30 seconds, then I'm able to refresh the page and use quickly before it does it again.  I shut the server down, cleared cmos, put in different memory, plugged it directly to my monitor where the server ran just fine with no hiccups. Moved it back to it's location and now here I am getting the disconnects again.

[FlyingTexan]

I hadn't noticed this error before as I hadn't gone to my profile yet. Hitting google now on this one.

[FlyingTexan]

I was hoping disabling that would make the difference but it has not. 

[FlyingTexan]

6 hours ago, itimpi said:

I believe that mover tends to process shares (and their contents) in alphabetical order..   There was probably not much spare space when it started on appdata and do nothing got moved.    You say it is currently moving data off the cache for the ‘homeserver’ share which is freeing up space?
 

The ‘appdata’ share seems to be bigger than the cache can hold which is not a good scenario to be in - any idea why it is that huge?

 

it is also worth mentioning that the ‘system share contents can normally only be moved if the Docker and VM services are stopped as these hold files open and open files cannot be moved.

I misread that. App data is roughly 5.5GB.  My system is still having this disconnect issue. It worked fine plugged into a monitor. Now it's disconnecting again.  The move has been completed and the SSD cache is as it should be. 

[FlyingTexan]

I'm still perplexed by the CPU showing no usage. Again when it was connected to the monitor all of those functions were normal. Before today I've never had this issue. 

[FlyingTexan]

Besides the diagnostic above is there a way to check networking settings I can post here for anyone that knows networking?

[JonathanM]

Your symptoms remind me of what happens when there are two devices on the network with the same IP. How are you managing IP assignment in general on your network?

[FlyingTexan]

5 minutes ago, jonathanm said:

Your symptoms remind me of what happens when there are two devices on the network with the same IP. How are you managing IP assignment in general on your network?

To be honest I don't really. I just have my desktop, server, and appletv, and linksys mesh network hardwired to a switch. The linksys mesh obviously handles all my wifi. The switch is plugged into my ATT gateway.  What should I be doing? How could i check if something is trying to use the same IP address?

[FlyingTexan]

Would a networking issue have anything to do with why the CPU shows no data but when plugged into a monitor and botting to the gui it does?

Edited June 24, 2021 by FlyingTexan

[FlyingTexan]

28 minutes ago, jonathanm said:

Your symptoms remind me of what happens when there are two devices on the network with the same IP. How are you managing IP assignment in general on your network?

I was trying to get My Server up and running and was in my ATT router doing IP passthrough.  The options I had were:

Allocation Mode:  Passthrough <selected>

Default server internal address: <address of server was put in>

Passthrough Mode: DHCPS-fixed <selected, also had dynamic available>

Passthrough Fixed Mac Address: <had address of tower listed which matched>

 

Then it asks for passthrough DHCP time.  I very rarely mess with any sort of networking configs so most of what I did was from google searches. 

[FlyingTexan]

I have removed the IP passthrough setup from the gateway for now. That appears to be the reason I'm hoping.  I don't know what I did wrong but the CPU data is there now and things are working.  I'll let it run for a bit and see if there are any other issues. If it's working in a couple hours I'll mark this solved and hope it can help someone else in the future!  Thank you for your help and comments that's what lead me to the ATT router IP passthrough settings!!!

[Arubial]

Enabling IP passthrough would send your public ATT IP address to the Unraid server. Basically your server was publicly exposed to the Internet. No wonder you were having issues. 

[JonathanM]

3 hours ago, Arubial said:

Enabling IP passthrough would send your public ATT IP address to the Unraid server. Basically your server was publicly exposed to the Internet. No wonder you were having issues. 

Yeah, I figured since the passthrough was no longer in place, no need to beat that horse until or if it came back to life.

 

ONLY expose individual ports that are strictly necessary for a known purpose, NEVER passthrough or DMZ your server.