Scale's UI freezes and becomes unavailable

[departy]

Hi yall,

I am with latest updates of TrueNAS Scale, I have been experiencing this issue since I use Scale. (Just a note I moved from Synology to TrueNAS Scale couple months ago)

For example I have a replication task, which I start and UI becomes unresponsive, when I hit refresh, i am stuck to this screen for uncertain amount of time. BUT SMB, NFS and other types of shares I use all work at full capacity!!! Just the UI

Since I am new to TrueNAS and Custom build NAS, here are the specs.

I run TrueNAS Scale in a VM on R720 server with ESXi.
4 x Intel(R) Xeon(R) CPU E5-2640
20 GB RAM ECC
No-cache discs so far
RAIDZ2 - 16TB total usable space (6x IronWolf PRO 4 TB)

Example tasks that this can happen:
Trigger Replication Task
Trigger Update App lists


If I can look at some logs, please let me know where exactly to look and will provide anything further needed.

Any help would be appreciated!

 

[morganL]

Is IP networking to the UI Ip address working?
Is SSH to the CLI working?

Can you detail the Replication task... I assume the replication tasks is not working either?
Both of your examples seem to be related to external IP address access?

You might want to test and review the IP networking set-up.. Is it one IP address on one "virtual port".?

 

[Kris Moore]

A debug file and a ticket on https://jira.ixsystems.com would be helpful. Chance that the middleware is crashing or getting so bogged down that the UI can't connect any more. We'll need to investigate.

 

[departy]

Is IP networking to the UI Ip address working?
Is SSH to the CLI working?

Can you detail the Replication task... I assume the replication tasks is not working either?
Both of your examples seem to be related to external IP address access?

You might want to test and review the IP networking set-up.. Is it one IP address on one "virtual port".?

IP network is working as mentioned above I have the access through SMB, FTP, NFS and etc on different VLANs and VPNs
SSH and CLI, havent checked, I never exposed the SSH, but will check it out once I have a chance.

Replication task I can confirm that starts WORKING as expected

 

[tcd]

Having a similar issue. This is accompanied by excessive CPU usage on the client running the browser. I suspect it is related to javascript event handling.

 

[morganL]

Having a similar issue. This is accompanied by excessive CPU usage on the client running the browser. I suspect it is related to javascript event handling.

The bug should be reported. If someone with the issue could run the latest nightly (this is a good approximation of SCAE 22.02.1), it would be helpful to know if the issue still exists.

 

[stavros-k]

The bug should be reported. If someone with the issue could run the latest nightly (this is a good approximation of SCAE 22.02.1), it would be helpful to know if the issue still exists.

Running on nightlies here, I don't have any replication tasks, but I can "confirm" that occasionally when clicking "Install" on an app (it doesn't matter if the questions.yaml is super small or super big), when scrolling (not immediately), it starts to freeze for some seconds and as you get to the bottom, it gets worse and some times it freezes completely. At the point where chrome just gives you a pop up that you either kill or wait the tab.

If you kill and refresh the same tab get's you right to the login screen and everything is snappy again. (even trying to install this same app).
If you "wait" the tab and try to refresh it does nothing.

I have it for sometime now, but it's totally random.
I tried to reproduce it, so I can make a ticket, but it's really random and not very often

To be fair, didn't check if there was any other background tasks happening at the time, and this test machine isn't that powerful..

If I find a reproducible way, I'll let you know.

 

[tcd]

It might not be totaly random it seems to happen when you choose an option that renders a pane with more information. Enabling ingress for instance.

Also I'm running an i9 with 64Gb RAM so it should be able handle the UI.

 

[morganL]

Running on nightlies here, I don't have any replication tasks, but I can "confirm" that occasionally when clicking "Install" on an app (it doesn't matter if the questions.yaml is super small or super big), when scrolling (not immediately), it starts to freeze for some seconds and as you get to the bottom, it gets worse and some times it freezes completely. At the point where chrome just gives you a pop up that you either kill or wait the tab.

If you kill and refresh the same tab get's you right to the login screen and everything is snappy again. (even trying to install this same app).
If you "wait" the tab and try to refresh it does nothing.

I have it for sometime now, but it's totally random.
I tried to reproduce it, so I can make a ticket, but it's really random and not very often

To be fair, didn't check if there was any other background tasks happening at the time, and this test machine isn't that powerful..

If I find a reproducible way, I'll let you know.

I'd suggest reporting a bug with whatever info you have... particularly what you were doing when it happened.

We might need to make an educated guess on which module might cause this type of behaviour and increase the testing focus. We can let you know if we think we have a fix.

 

[stavros-k]

I'd suggest reporting a bug with whatever info you have... particularly what you were doing when it happened.

We might need to make an educated guess on which module might cause this type of behaviour and increase the testing focus. We can let you know if we think we have a fix.

Here it is, managed to find an app that "should" be reproducible

https://jira.ixsystems.com/browse/NAS-115796