Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

nicho

macrumors 601
Feb 15, 2008
4,250
3,250
The OP destroyed his Apple. If he'd removed the Promise software at the first sign of trouble it would have been fixed.

The OP destroyed his Apple. A self-inflicted wound.

A fair title might be

Warning: Having Promise Utility installed led me to complete destroy my mac

Perhaps it's actually PSA worthy :)
 
  • Like
Reactions: OkiRun

Slash-2CPU

macrumors 6502
Dec 14, 2016
404
268
The OP destroyed his Apple. If he'd removed the Promise software at the first sign of trouble it would have been fixed.

The OP destroyed his Apple. A self-inflicted wound.

It's a utility that's offered by the vendor for configuring the RAID system inside the Mac Pro that hosed his files. Hardly self-inflicted. Par for the course in your replies, it's never Apple's fault. Do you own stock or what?

Very glad I held off on buying a 7,1 for things like this. Pay $6k+ to test Apple's hardware for them. Nah, I'll wait.
 

Snow Tiger

macrumors 6502a
Dec 18, 2019
854
634
It's a utility that's offered by the vendor for configuring the RAID system inside the Mac Pro that hosed his files. Hardly self-inflicted. Par for the course in your replies, it's never Apple's fault. Do you own stock or what?

Very glad I held off on buying a 7,1 for things like this. Pay $6k+ to test Apple's hardware for them. Nah, I'll wait.

The MP7,1 is a beautiful machine , but the early adopters are the pioneers who get the arrows in the back .

For production rigs , perhaps waiting a little longer for the OS and apps to mesh together properly on a brand new platform would be a good piece of advice .

Darn thing is , the MP7,1 was really long overdue by some seven years and a lot of ppl couldn't wait any longer if they wanted to remain in the macOS world .
 

ZombiePhysicist

Suspended
May 22, 2014
2,884
2,794
Yea I’m not sure why we are blaming the victim here. The software is garbage. It destroys data. The psa is valuable to all of us. Thanks to the op for sharing.

Should the original op had more backups, sure, most of us should. Should the original op somehow have known the software was bad and pulled it sooner, who knows, that assertion is employing a lot of hindsight and these things are sometimes hard to pin down and diagnose.

But ultimately op didn’t write the crappy data destroying software and installing it wasnt unreasonable.

Agree with @Snow Tiger that a lot of us couldn’t wait. Some of us pioneered implmentations and shared howtos, sharing many obstacles and mishaps along the way. We should be thankful for the free information exchange. I for one, am. Ymmv
 
Last edited:

thekev

macrumors 604
Aug 5, 2010
7,005
3,343
Ok I can try to do this at a later point when these projects are over, right now I cant clean install AGAIN.
I did that last night and didn't migrate any user settings...

I'm not a big fan of people suggesting that anyone reinstall over and over again. It's typically very unhelpful if it doesn't work the first time. If you depend on this machine, and you don't have backup hardware that can be used in a pinch, then you really need a known stable configuration of some kind. This means with the OS, applications and any third party drivers at a point where they are known to be working without further updates installed until you once again have a stable configuration.

Time machine backups and stuff like that can still fail. Time Machine is a very stupid backup system in general. Apple should have done better there.

Really though, it's always possible to encounter something like this, and just randomly reinstalling stuff as people suggest is unlikely to be of any help or worth the time involved.

Good luck with your deadlines.
 
  • Like
Reactions: chfilm

chfilm

macrumors 68040
Original poster
Nov 15, 2012
3,430
2,116
Berlin
Yea I’m not sure why we are blaming the victim here. The software is garbage. It destroys data. The psa is valuable to all of us. Thanks to the op for sharing.

Should the original op had more backups, sure, most of us should. Should the original op somehow have known the software was bad and pulled it sooner, who knows, that assertion is employing a lot of hindsight and these things are sometimes hard to pin down and diagnose.

But ultimately op didn’t write the crappy data destroying software and installing it wasnt unreasonable.

Agree with @Snow Tiger that a lot of us couldn’t wait. Some of us pioneered implmentations and shared howtos, sharing many obstacles and mishaps along the way. We should be thankful for the free information exchange. I for one, am. Ymmv

Well thank you.

People here are suggesting to notbe an early adopter, to wait till the software has been through its growing pains etc... well that’s why you install updates right? To get rid of the growing pains that plagued the initial versions, and that’s why I had to install this promise tool, to fix an issue I had with this raid because the promise support couldn’t figure it out without a log file, which the old version was unable to generate under Catalina.
This had nothing to do with the Mac Pro being a new device actually, but with promise messing up their driver somehow and adobe software being EXTREMELY nervous apparently. FCP and Resolve where unaffected by this, but of course adobe’s entire graphics kernel crashes at the first sign of trouble ;)
 

tommy chen

macrumors 6502a
Oct 1, 2018
907
390
videoerror:

this should not normally cause the problem !?!
 

Attachments

  • Screenshot 2020-02-12 at 23.41.18.jpg
    Screenshot 2020-02-12 at 23.41.18.jpg
    30.9 KB · Views: 135

Adult80HD

macrumors 6502a
Nov 19, 2019
701
837
BTW, this is not the first time that the Pegasus software has been able to hose up a machine. I had something similar happen with either Yosemite or El Capitan. Ever since then I've been loathe to ever install their software.
 
  • Like
Reactions: chfilm

Slash-2CPU

macrumors 6502
Dec 14, 2016
404
268
Having a hardware RAID controller on a workstation with that much available CPU power is a bit nonsensical. Whole lot of additional complexity and cost for no real extra speed or reliability increase. Doing RAID 5 calculations at 1+GB/s for even an 8-core CPU is very, very little load.

Given Apple's history of long term support for its RAID cards(or lack thereof), I don't see the appeal. That you need to use the hardware RAID controller just to have 4 drives worth of bulk storage is a shame.

Mac Mini plus 2-3 TB3 enclosures is looking more attractive every time I see posts like this. TB3 makes the Mac Mini itself a module.
 
  • Like
Reactions: bsbeamer

defjam

macrumors 6502a
Sep 15, 2019
795
735
Having a hardware RAID controller on a workstation with that much available CPU power is a bit nonsensical. Whole lot of additional complexity and cost for no real extra speed or reliability increase. Doing RAID 5 calculations at 1+GB/s for even an 8-core CPU is very, very little load.

Given Apple's history of long term support for its RAID cards(or lack thereof), I don't see the appeal. That you need to use the hardware RAID controller just to have 4 drives worth of bulk storage is a shame.

Mac Mini plus 2-3 TB3 enclosures is looking more attractive every time I see posts like this. TB3 makes the Mac Mini itself a module.
RAID calculations have been trivial for decades and is not an area of performance concern. The largest impact for RAID is a result of the additional I/O necessary used by RAID.
[automerge]1581557654[/automerge]
What is nonsensical is to use software RAID on any system, except maybe for RAID-0 scratch volumes.

Hardware parity RAID with persistent writeback cache is the only way to go.
Hardware RAID does not permit a system to spread the RAID construction across multiple controllers. Software RAID does.
 

Slash-2CPU

macrumors 6502
Dec 14, 2016
404
268
What is nonsensical is to use software RAID on any system, except maybe for RAID-0 scratch volumes.

Hardware parity RAID with persistent writeback cache is the only way to go.

Then why does my ixMicro High Availability TrueNAS ZFS system not have any hardware RAID controllers? The whole industry is moving away, if not already, from having logic in the controllers.

In event of controller(or in my case system failure), I can move all the drives over to the spare chassis, press power button, and it’s all back in a few minutes.

Same for Apple or SoftRAID software RAID. We have enclosures that are RAID 0 and RAID 5 that get physically moved between systems with no reconfigure.

Bet if the Promise controller flakes out, you’re waiting for a call back during business hours. That’s AFTER you receive the replacement.

The real problem is APFS not supporting RAID, so Apple had to band-aid it. To be fair, we’re asking too much of APFS. It’s designed for iPhones and watches, not desktops. Too bad no ZFS support, like real workstations have. ZFS slog ZIL is a persistent write back cache.
 
Last edited:

AidenShaw

macrumors P6
Feb 8, 2003
18,667
4,677
The Peninsula
Then why does my ixMicro High Availability TrueNAS ZFS system not have any hardware RAID controllers?
That is a hardware NAS appliance, not software RAID running on the host OS. https://www.ixsystems.com/truenas/

If the base filesystem is ZFS Stripe/ZFS Mirror/RAIDZ/RAIDZ2/RAIDZ3, it is funtionally a hardware RAID controller (especially if using iSCSI/FC for block protocols).
 
Last edited:
  • Like
Reactions: chfilm

26139

Suspended
Dec 27, 2003
4,315
377
I don't want to start a Windows versus Mac debate with this post but rather just say I would like everyone to keep this thread in mind when labelling Windows as problematic. Issues happen on all platforms and, as the original post shows, Macs are not immune.

Brrrrruuuuuhhhhhhh.
 

Slash-2CPU

macrumors 6502
Dec 14, 2016
404
268
That is a hardware NAS appliance, not software RAID running on the host OS. https://www.ixsystems.com/truenas/

If the base filesystem is ZFS Stripe/ZFS Mirror/RAIDZ/RAIDZ2/RAIDZ3, it is funtionally a hardware RAID controller (especially if using iSCSI/FC for block protocols).

There is no hardware RAID controller in the box. It’s a SuperMicro Xeon box with a different plastic front grille. All software. Open-source Linux OS running open-source ZFS RAID software. For that matter, all the little linux-based SoHo NAS boxes like Drobo or Qnap have no hardware RAID controllers. They’re powered by some consumer-grade CPU, usually even without ECC RAM. They’re all softRAID. OS with CPU running RAID is software RAID.

A core concept of ZFS is to be software-defined and controlled with an x86-64 CPU.

Being functionally a RAID controller while not having a hardware RAID controller is the very definition of software RAID.

OpenZFS could have been added so beautifully to 10.15. Drop in 4x12TB platter drives and 4x2TB NVMe SSD’s. Set up the NVMe as cache. If you really want to see it scream, add a 900-series Optane drive as the SLOG. ZFS also loves RAM. The 7,1 is pretty close to an ideal ZFS workstation, except it doesn’t support ZFS.
 
Last edited:
  • Like
Reactions: Schismz

Schismz

macrumors 6502
Sep 4, 2010
343
395
[...]

The 7,1 is pretty close to an ideal ZFS workstation, except it doesn’t support ZFS.
Well... ZFS kinda, sorta, mostly existed on OS X for a few minutes before the illusion flickered out heralding the triumphant further entanglement of HFS+ ... and now we have APFS, which mostly isn't there yet <seasons change, years go by>.


Bummer. Sun did a lot of cool things before The End.
 
Last edited:
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.