System unreachable for several minutes.

[vaschthestampede]

Twice today the system was unreachable for several minutes.
I report the screen and journalctl of those minutes.

Can you give me information on what causes them and how to prevent them from happening again?

 

[NugentS]

The problem is caused by a lack of hardware.

Seriously - please read the forum rules and post your hardware

 

[HoneyBadger]

This is very likely related to the deduplication issue we're tracking in your other thread.

 

[NugentS]

That seems very likley

 

[vaschthestampede]

Seriously - please read the forum rules and post your hardware

Sorry about that!
CPU: 2 x Intel(R) Xeon(R) Silver 4310 CPU
RAM: 512GB ECC (8 x 64 GB)
Data VDEVs 9 x RAIDZ2 | 11 wide | 20.01 TiB
Dedup VDEVs 4 x MIRROR | 2wide | 3.49 TiB

This is very likely related to the deduplication issue

That's what I thought too but in the previous blocks the whole system didn't block.
In these last two blocks not even the WebGUI was reachable.
It could be a worsening of the situation.

 

[NugentS]

From the other thread - I imagine that the issue is getting worse as your DDT is getting / has got too big for the dedupe vdev amd ARC.

You need to fix (if they are fixable) those issues first

 

[vaschthestampede]

I don't think it's a worsening of the previous problem.
Because I had to reboot and rebooting gives this error:

You need to fix (if they are fixable) those issues first

We are taking two more disks for dedup.

 

[NugentS]

2 more disks will help (a bit) but not solve the issue. If the DDT's have been written to the HDD's because you over extended on the metadata vDev then you have a problem which will remain.

 

[HoneyBadger]

Because I had to reboot and rebooting gives this error

Did you reboot the TrueNAS system because it was no longer responsive? This is likely because the underlying disks are busy with the I/O for deduplication.

As mentioned in the other thread, I'd like to see an additional zpool status -D and zpool iostat -v if possible during the times of non-responsiveness.

 

[vaschthestampede]

Even with a fresh installation of TrueNAS I get this error.

It is currently in the import phase.
What does the middleware process do?