Hi all,
UPDATE: I closed the post (the timebox I gave myself to understand the issue is now over). Thank you all for the help ^^
DISCLAIMER: The objective of this post is to understand how people would debug issues like these when real data is involved and get to the bottom of the problem. The objective is NOT to "restore service" but to understand what failed. The tone of the post is voluntarily not serious to keep it light.
I am playing a little with TrueNas Scale and ZFS. I was trying to use a second NVME disk via USB to do a replication once a day of the main pool, however I had issues with this secondary pool being SUSPENDED for "too many errors". This pool is not directly write/read by users/apps, but it is just there to be "replicated on" once a day.
Now, please, I know that using disks via USB is not advised. Also I am not interested in recovering the data, since there is nothing real on it. What I am doing is testing to see if the system is brittle, and if it is, how to debug if there is a real issue.
Now to the point. The pool is SUSPENDED. Good. Why? I mean, the real reason why. To see if the system can be used in real life it needs to be debuggable.
Let's start. The pool is SUSPENDED:
pool: tank-02
state: SUSPENDED
status: One or more devices are faulted in response to IO failures.
action: Make sure the affected devices are connected, then run 'zpool clear'.
see: https://openzfs.github.io/openzfs-docs/msg/ZFS-8000-JQ
config:
NAME STATE READ WRITE CKSUM
tank-02 UNAVAIL 0 0 0 insufficient replicas
xxx-xxx-xxx-xxx-xxx FAULTED 3 0 0 too many errors
errors: 4 data errors, use '-v' for a list
To which you may ask: why? Too many errors (the -v says nothing more). Well that doesn't help, does it. When you run zpool clear:
# zpool clear tank-02
cannot clear errors for tank-02: I/O error
Incredibly useful as you can see. dmesg to the rescue?
WARNING: Pool 'tank-02' has encountered an uncorrectable I/O failure and has been suspended.
Thanks? I guess. I know it it trying to safeguard data but again... why?
Before you ask:
- SMART checks are good
- Yes, I restarted the device. As soon as you try to use/mount/import you get to the same issues.
- Nothing else peculiar in dmesg. I mean the USB was
usb 2-4: USB disconnect, device number 12whatever the reason why. I mean, kick me if I know why TrueNas scale decided that having/sys/module/usbcore/parameters/autosuspendto2is a good idea but again, that is not the point. I need ZFS to reply to me what is the issue for its point of view.
I have read a lot online. Maybe it is the temperarure (usb enclosure heating up), maybe it is the cable, power, "it is the usb controller", or the chipset doing the usb -> nvme... However, therey are not saying what to check. People is guessing. I saw more tech behind reading tea leaves.
My question for you all is this: ZFS SUSPENDED one of my pools. It (seems to me) is refusing to fix it. Refusing to do anything with it and to tell me why. So, in a real world case, how to debug it? If I have to trust my data to it, I don’t want the only option to be “use many disks and just replace one and the cable when ZFS poo-poo”.
How to know the cause?
Thank you for the help.
PS: I am sure I am missing some very basic ZFS knoweldge on the topic, so please let me know what else can I do to make ZFS talk to me.
Hi,
Thanks for sharing. I agree with you 100% and I think everybody commenting here does. The whole point of the thread however was to understand if/how you can identify the location of the problem without guessing. The reality is I got to the conclusion that people... don't. Like you said people know ZFS is fussy about how does he speaks with the disks and the minimum issue it has it throws a tantrum. So people just switch things until they work (or buy expensive motherboards with many ports). I don't like the idea of not knowing "why", so I will just add on my notes that for my specific usecase I cannot trust ZFS + OS (TrueNas scale) to use the USB disk for backups via ZFS send/recieve.
I would like to add that I am not trying to
mirrormy main disk with a usb one. I just wanted to copy the zfs snapshots on the usb drive once a day at midnight. ZFS is just (don't throw stones at me for this, it is just my opinon) too brittle to use it this way too. I mean when I am trying toclean/recover the pool it just refuses (and there is no one writing on it).In my case there was no switching however. It was a single nvme drive in a single usb line in an enclusure. It was a separate stripe to just recieve data once a day.
Not without good logs or debugging tools.
I decided I cannot trust it so unfortunately I will take the usb enclosure with the nvme, format it with etx4 and use Kopia to backup the datasets there once a day. It is not what I wanted but it is the best I can get for now.
About better solutions for the my play-NAS in general, I am constrained with the ports I have. I (again personal choice - I understand people disagree with this) don't want to go SATA. Unfortunately, since I could not find any PCIe switch with ASM2812I (https://www.asmedia.com.tw/product/866yq74SPBqRdtgC/7c5YQ79xz8urEGr1) I am unable to get more from my m2 nvme pcie 3x4 (speed loss for me is not an issue, my main bottleneck is the network). It is interesting how you can find many more interesting attempt at it in the PIs ecosystem but not for mini PCs.