mfi0: COMMAND 0xfffffe0000ef TIMOUT AFTER on Dell R730xd

[Alan McKay]

Thanks for the detailed response. Is there a list of recommended hardware for FreeNAS? Section 1.2.3 of the manual (for 9.3 at least) mentions only the Fusion I/O card, and then points to the FreeBSD list which I referenced earlier.

I think at this point my best option - which isn't really what I want to be doing - is to move data over to another system on a temporary basis, and spend some time playing with this hardware in the lab. What I might end up having to do is use the native RAID card and reinstall FreeNAS without the juicy benefits of ZFS. Which would kill me because the whole reason I went with FreeNAS was for the benefits of ZFS. But obviously I did not do the correct amount of homework.

I don't have a response back yet from Dell on whether or not this PERC is a separate card that I can replace. If so, I will look at doing that. Get an HBA and try to do it right.

 

[jgreco]

Thanks for the detailed response. Is there a list of recommended hardware for FreeNAS?

http://www.freenas.org/hardware-requirements/

There's several posts in this subforum that discuss hardware suggestions.

What I might end up having to do is use the native RAID card and reinstall FreeNAS without the juicy benefits of ZFS.

FreeNAS no longer supports UFS; ZFS is your only option.

But obviously I did not do the correct amount of homework. I don't have a response back yet from Dell on whether or not this PERC is a separate card that I can replace. If so, I will look at doing that. Get an HBA and try to do it right.

And now you know why it kills me to reply to these threads. :-/ I hate to be the bearer of not-great news to people who have good intentions.

 

[Alan McKay]

The hardware requirements page just points back to the FreeBSD page again - I'll try to fish stuff out of the forums.

I don't mind compiling a summary if it will benefit others in the future.

I also see that iX now has 1U and 2U units. I approached them for a quote some time ago and the smallest they had at the time was 3U

I just went back to them for a quote.

I originally wanted to use Supermicro gear - what's the general feeling on it?

 

[Ericloewe]

I originally wanted to use Supermicro gear - what's the general feeling on it?

Extremely popular and it tends to work.

 

[pirateghost]

The hardware requirements page just points back to the FreeBSD page again - I'll try to fish stuff out of the forums.

I don't mind compiling a summary if it will benefit others in the future.

I also see that iX now has 1U and 2U units. I approached them for a quote some time ago and the smallest they had at the time was 3U

I just went back to them for a quote.

I originally wanted to use Supermicro gear - what's the general feeling on it?

https://forums.freenas.org/index.php?threads/hardware-recommendations-read-this-first.23069/

It's a sticky in this very forum

 

[Alan McKay]

Hey that's some good reading right there! Thanks.

 

[Alan McKay]

Update on this ... we popped out all the drives and reset them in the drive bays and the problem went away. That does not give me a great feeling about this hardware to be honest. But at least I am finally able to re-import my ZFS array and I have confirmed that my data is still there!

However I am scheduling some changes as recommended in the other thread I bumped the other day with details on the R730xd. I have to set up a 2nd FreeNAS box, copy my data out, reconfigure the R730xd, then copy it back in.

 

[Alan McKay]

Another update - the system operated for a few hours after reseating the drives, then failed again.
But finally there is a HW error reported in the Dell iDRAC - ECC memory error on the RAID card's cache.
So Dell is coming in to replace the RAID card.

It took over a week after the failure for the error to report itself in the iDRAC error log.

I had a feeling it was bad RAM somewhere based on the symptoms.

 

[Robert Trevellyan]

So Dell is coming in to replace the RAID card.

 

[Alan McKay]

Yeah I've never been so happy to see a HW error

 

[Mirfster]

You do realize that Dell farms out their Service Calls and 9 out of 10 times the people that arrive I would tell to simply hand me over the card instead of letting them touch any of my systems? ;)

 

[jgreco]

Another update - the system operated for a few hours after reseating the drives, then failed again.
But finally there is a HW error reported in the Dell iDRAC - ECC memory error on the RAID card's cache.
So Dell is coming in to replace the RAID card.

It took over a week after the failure for the error to report itself in the iDRAC error log.

I had a feeling it was bad RAM somewhere based on the symptoms.

You should see if you can get them to replace the card with an HBA.

 

[Alan McKay]

I can't do that since it might risk losing all my data. These cards do have HBA mode and from the other thread about these systems it seems it works fine. So my plan is to migrate my data to another system, convert this to HBA mode, then migrate it back.

 

[jgreco]

I can't do that since it might risk losing all my data. These cards do have HBA mode and from the other thread about these systems it seems it works fine. So my plan is to migrate my data to another system, convert this to HBA mode, then migrate it back.

No, they don't have an HBA mode, and they cannot be made to work sufficiently to the task. I refer you to the very first reply in this thread: "JBOD mode is insufficient."

You can, of course, choose to chance it and pray that there aren't hard drive problems being obscured by the MFI driver and firmware, which seems like a perfectly reasonable thing to do, but the fact of the matter is that I thought that at one point as well, but in the years since, I've seen too many strange problems pop up for users who actually do this, up to and including the loss of pool, to feel comfortable suggesting anybody actually do this.