Mac mini M1 - Not able to mount LaCie 5Big Thunderbolt storage

April 8 - 2021. 5 few days ago I received a notice that LaCie had issued an update to the LaCie RAID manger that would work under Big Sur 11.2.3. Today, April 8th, I set aside the afternoon to download and install the RAID driver and physically move my 5BIG from my Intel Mac Pro 6,1 (2013) Thunderbolt-2 to my Mac Mini M1 (16GB ram/2TB storage). So this is LaCie Raid Manager 2.9.3.207. This time the install under Big Sur went fine - I had to approve the software's source with Privacy as expected and this time it went okay. The drive was recognized and I could read and write to it. However, it's performance is seriously impacted. On my Intel MacPro (TrashCan 6,1 Thunderbolt 2) I get at least, 400/400 MB/sec by any measure (transfer a really big video file and stopwatch it, Blackmagic, etc) - but on moving the drive to the M1 my speeds are roughly write 55MB/sec; read 270 MB/sec...there is some variability but this is the essence of it - not much better than a USB2 drive and on par with a FW 800.

So this is a serious problem. This drive, while readable, is useless for ay serious photographic/video work.

I spent some hours on the phone with tech support today and they do not have an answer for this. I swapped every TB2 cable and Apple adapter I have* - so it's not a cable or adapter issue. It's a driver issue - entangled with caches/buffers/optimization and all that.

*I have 3 Apple TB 2<-->3 dongles I have 2M and half-metre cables from different manufacturers - I counter-swapped, and all these work fine on the TrashCan - and when connected on the clearest/simplest path on the M1 Mac - much less performance.

Two questions

1. what numbers are you seeing*
2. How do we mount a claim on this.

*the reason I am asking this is if your 5Big (whatever) is performing differently than mine - I have done 2 failed drive swaps over the years and although I have replaced with IronWolf, maybe I have introduced something incompatible or less optimized. Nonetheless, the LaCie Techs cannot discern this and the drive currently does do 400+/400+ on the Intel and 55/270 on the M1.

I do not relish dumping this drive system - but it is useless on the new future of Macs. The LaCie tech sort-of admitted to this but has no solution and so we must apply some pressure.

Nest Steps. BTW - I am sure this applies to 6Big and 12Big as well....so this is not a small community that has had their hardware back-watered very prematurely.

jc - April 8-2021

g'morning Hendrik,

Tuesday - Apr 27 - 1pm Toronto time.

I upgraded this morning to BS 11.3 on my MacMini. The good news is that I now have a fully operational OWC ThunderBlade (1TB, 4x256 ssd) running with Softraid 6.03 on the M1 at roughly 1400/1400 ... which is about what I'd expect - using SoftRaid 6.03. So, that is resolved.

I did a LaCie 5BIG test before and after the OS Update, just in case Apple had mysteriously patched something in BS 11.3 that affected the kernel level IO - but sadly it performed the same after - showing the kind of numbers we have all been seeing all along.

Referring to the handshake issue at the hardware level, it is interesting to note that SoftRaid is now running acceptably. I do have some very early screen shots showing 1800/1600 on my blade when I got it 2+ years ago on my 2017 15" MacBookPro - (which is/was a native TB-3 machine) and the speed has degraded a bit since then but I suspect that may have something to do with the SSD's in the ThunderBlade - and I've had much discussion with OWC on this. I did finally completely rebuild this in the fall on my Intel machine when the speed had dropped to the 300's and refreshing it and re-populating it has brought it back to the 1400's I am running now on both my Intel and M1. On the M1, it is running on a Thunderbolt3 chain and the other device on the chain is a LaCie 10TB single-drive 7200rpm disk.

My point is that THEY (OWC) have made this work - so if there is some sort of handshake or timing issue, then they have been able to circumvent it; or there is no handshake issue and it's just a programming issue in the LaCie driver (which is what I really suspect). I've never experienced a speed problem with the LaCie 10TB on the M1 - and it's not raid so it has no need for the LaCie raid software.

I'll be writing back to my contact on my case at LaCie later today and will fill them in. This is their response to me last week.

"Hello Jim Camelford,

Thank you for your email response and for sharing that video.

First of all, I apologize for the delay in answering. The video was already shared with the Subject Matter Expert team, alongside a few Apple forums where people shared their issues, and they will look deeper into it. I thank you for taking the time for making that video, helping fix the performance issue faster for all the customers.

Now we have to wait for the next upgrades in the software, that will improve compatibility and thus, the performance."

I will be pointing out the above and asking for a progress report and will let you know.

On the issue of you having to create a partition and do a fresh load of BS-what a waste of your time. We know the problem is real.

Do you have a case open with LaCie ?

jc

Actually this is a LaCie issue that is dragging on longer than it should. Compare LaCie/Seagate’s and G-Technology/WD’s response to the problem vs Promise Technology’s response. Promise Technology has already release a downloadable fix by the end of Dec 2020. LaCie/Seagate can only add to their support page the follow notice. Not compatible!!! While G-Tech/WD is pretty much ignoring the problem until the general public figures a work around.

My take is that these companies didn’t work hard enough to get their systems aligned with Apple’s M1 and Big Sur.

g'morning, Hendrik

Yesterday I wrote a formal letter to my case contact at LaCie requesting it be forwarded to their management. It included a technical case update as well as strategic positioning; a request for more transparency on the issue; a commitment to fix it and direct contact to me from their senior technical support management.

My situation was also featured on "The Constant Geekery Podcast" out of Britain, yesterday. It was discussed; This podcast has only about 1000 subscribers and it's worth noting that there wasn't much feedback from them - despite they're being people keenly interested in new technology, review, etc.

And I reviewed as much as I could, posts on the Internet regarding this issue with LaCie. There was a small flurry of activity early on - but most of the posts have been dormant since December. In any event, the audience is pretty small.

Which leads me to conclude there isn't much interest in this topic and the number of people affected now is relatively small. Maybe it will grow this year as M1's or their successors become the dominant strain of Macs. Nonetheless, LaCie is aware of it and the likely implications.

For even this thread ... right now ... is it just you and me ?

While we could work on mounting a wider awareness campaign, I am thinking it won't amount to much. If we did, my approach would be to contact 9-5 as you suggest and other YouTube channels (I have several in mind) ..... I can certainly adapt my narrative to a few paragraphs. As a question, I'd be asking for some speed reporting feedback since the connectivity issue is resolved (see my SoftRaid comment at the end, however...)

However, for now I'd like to see how far I get with LaCie - a sugar as compared to a vinegar approach. There are some Apple implications - but my experience with both the LaCie and SoftRaid issues is that it IS up to the 3rd parties to fix the problem with their software, working in conjunction with Apple to ensure the right driver API's are exposed on MacOS to make this work. All these players, in the end, have a vested interest in making this work. I own lots of LaCie products, my 5Big is the most expensive, though....and if I have to look elsewhere to fix this and LaCie hasn't come to the table, then the implications of THAT are pretty clear. Anybody else with LaCie RAID storage is eventually going to experience this - although as I've said, the problem is not well known.

I'd like to wait a week and judge LaCie's response.

My 5Big is attached to my MM M1 - it's 5 drives (5 x 4TB, ea) forming one RAID 5 group - and it's working for now so I can use the 5BIG as a backup - at least this is faster than re-attaching it to my Intel Mac and accessing it as NAS - that did provide sustained writes/reads of about 100MB/sec .... but backing up small files just killed the speed - so all-in-all I'm more practical with it being crippled, but locally attached for now.

As for the SoftRaid issue - with my OWC ThunderBlade RAID.....SoftRaid needs Big Sur 11.3 (which is in public beta release 7) and I am on 11.2.3 ..... I have been assured that the latest SoftRaid works on the public beta and so when 11.3 is finally released that should resolve the ThunderBlade issue. I did have it working for a short while on BS 11.1 but I moved to 11.2 as quickly as I could to mitigate the Bluetooth connection issues with my magic mouse. In 11.2 Apple, according to SoftRaid, actually accidentally removed the piece of enabling MacOs code needed for the driver - so it never worked in 11, fixed in 11.1, broken again in 11.2 and is apparently working in 11.3.....so I do have confidence in that.

My confidence in LaCie is lower - but their risk from the community at large is much bigger and I am hoping that, in conjunction with our continuing to update them will move them to resolve this.

jc 10am Toronto, April 13

Good evening, everybody.

I am writing this on April 28th, About 930pm Toronto time

I have sent some eMails to my contact at LaCie and I have not received responses - so nothing to report on that.

Tonight, however I decided I would try a little experiment-at least it was kind of fun.

I wanted to investigate the issue of whether the Thunderbolt "3 to 2" adapter out of the M1 was causing an issue or if it had to do with some of the handshaking in the hardware, as was previously mentioned.

On my M1 MM I have an Apple-Branded 3-2 adapter running out of the back of the M1 to a Thunderbolt 2 chain comprising my LaCie 5Big (configured as a RAID 5, 20TB ie 5x4=20 with 16TB user space), chained to an OWC "Toaster Dock" that can host 2 bare-bones SATA drives and from that to my DELL U3011 secondary monitor using a mini-display port cable to run that monitor at 2560x1600.

The monitor works perfectly.

And, as you know, the speeds we are getting on the 5Big are in the vicinity of 55/260MBS give-or-take (I've seen BlackMagics with worse...)

So, for fun, I put two fairly old 1TB Baracuda drives into the Toaster Dock - they were independently configured as HFS+ and I tested their speeds using BlackMagic (3GB Stress) - both running at 113/119 MBS write-read....which is what I would reasonably expect.

Then I used Disk Utility to group these into a 2-Disk RAID 0 and re-ran BlackMagic. As expected, performance increased, but only slightly, to write-151 / read-138 MBS. I have to admit I was a bit surprised that in all the tests the write speed was slightly better than the read speed. Speeds didn't nearly double, as I had hoped, but did increase.

So, now I have a "cheap and cheerful" RAID 0 running with Apple RAID on 2 quite ancient 1TB hard drives. And it is writing at least twice as fast as my LaCie 5Big RAID5. Not sure I'd put anything really critical on this, though !!!! And, remember, it's only 2 drives so the benefit of striped reads hasn't really kicked in on this test case to beat the 5BIG's 270.

Then I brought OWC's SoftRaid 6.03 into the picture. I let it take over the drives and re-booted the Mac. Speeds were virtually identical - about 1 MB/sec apart - so not significant.

So.....here's my point......I think the bus and handshaking are running superficially okay. That's about the speed I would expect from these drives - so I do not believe this is an inherent error in the hardware.

I re-tested my 5BIG right after this and once again got numbers in the 55 / 250 MBS range.

So, if my 'cheap and cheerful' RAID 0 setup using an OWC Toaster Dock can best my 5-disk, Lacie 5BIG RAID-5 then I surmise it is a fixable software problem in LaCie's drivers due to bad coding. At least that's what I hope; and not some never-to-be-addressed in Apple's hardware bus.

I'd love to see more BlackMagic tests from people posted to this thread, as well as some impact statements from people impacted by this situation that has dragged on now for more than 5+ months with very little comment or commitment from LaCie/Seagate.

My impact statement is that this has ruined my investment in this 5Big hardware, and I will need to look elsewhere.

With that we can continue to pressure LaCie to fix this.

With that being said, today I began to investigate purchasing a new RAID - this has to get fixed. I can't work with 55/270. At this point my shortlist did NOT include LaCie.

What about you ?

Hello all,

I'm facing exactly the same issue with the same setup (MacMini M1 as well as LaCie 5Big TB2 with an Apple adapter to TB3).

I guess we have to wait for an update from LaCie/Seagate to get this fixed. Unfortunately, their support is quite a mess.. I've opened two tickets (Mid Nov and around 12/01) with no answer or even a response yet.

I'm glad for every help and support the internet can provide ;)

Stay healthy

Hey ... hope the stars were awesome tonight.

I got some feedback from LaCie late on my day, today. We are scheduled to speak on the phone on Thursday about 4pm Toronto time - so we will have a discussion and see how that goes.

The gist of their eMail back to me today was that the .207 release was 'interim' and they were looking for field input - and with that can hopefully improve the performance; and learn if this is a one-off problem or a more wide-spread problem ! I would have thought some lab testing in this matter would have been pretty self-evident and they wouldn't have needed field input to detect that it was running at about one-eighth of its capable speed.

However, I have/and will commit to work with them to best advantage to test whatever; and provide any useful data I can.

But, if we can chat 'offline', that might also be beneficial.

In any event, I will report back with Thursday's conversation.

jc