AAE-9173 - Ran Out Of CPU Power

[Darryl Ramm]

Do you have any external drives? Was the Baraccuda or Evo in an external drive carrier/dock/chassis? I'm talking USB or Firewire or maybe Thunderbolt expansion chassis etc.

There are lots of people having CPU errors, so far nobody else has reported your suggestion works for them, hopefully it does and they will speak up, but I would not be surprised that there are many more issues that different people are facing.

All the utility program does you are using does is periodically poking the drives, the only drives that utility program is supposed to help with are external drives that ignore intended ways of things working and put themselves to sleep/suspend early on. Internal SATA and PCIe drives usually follow intended behavior properly (the behavior/specs there are clearer to manufacturers to follow).

That's what the utility program you are using is designed to do, what the author of that program wrote it for. It's absolutely possible if you have an external drive that is putting itself to sleep improperly then you've stumbled on a fix here that would be simple, and useful, to understand. If there is no external drive involved here then things will be more difficult to explain, that is absolutely not saying it's not working, I'm just curious to understand what is going on. (I have a technical background working in storage technology areas).

Going from SSD to HDD back to SSD and seeing more CPU errors on the HDD seems to show you are on the right path here. That was great stuff.

[lesbrunn]

Originally Posted by Darryl Ramm

There are lots of people having CPU errors, so far nobody else has reported your suggestion works for them, hopefully it does and they will speak up, but I would not be surprised that there are many more issues that different people are facing.

+1. While sleeping drives could definitely be a cause, there are many other factors (hardware, software, workflow, system configuration /optimization etc) that could potentially cause these issues.

In my Windows power settings, everything including drives and screens are set to never sleep. I do have a couple of external drives, and even though I don't use them for current sessions, I agree they could be a cause by demanding CPU attention when PT is running. But that's one of the reasons why I have the first core set aside for OS use. I'll have to look for a Windows 'drive-poker', and see if that helps.

[The Drummer]

Let me be clearer:
My System:
Mac Pro(Mid 2012)
2 x 3.46 GHz 6-Core Intel Xeon
64 GB 1333 MHz DDR3
NVIDIA GeForce GTX 970 4096 MB
Sierra OS 10.12.6

PT 12.8.2.106/2x 16x16 Avid HD interfaces

The Mac is used both for Audio(PT) and for Video (Adobe Creative Cloud Suite)
I have two separate System Drives(Audio & Video), both Samsung Evo 1TB SSDs
both running Mac OS 10.12.6, both in the Mac Pro’s internal trays.
OtherAttached Drives:
PCIe card with two 1TB Samsung Evo SSDs in a Raid 0 config as the Video files drive
Crucial 1TB SSD as a Video Cache drive and VI sample set drive(Video cache and Audio sample sets never accessed at the same time)
3rd Mac Pro internal tray: Samsung Evo 525GB as 2nd VI sample set drive.
4th Mac Pro internal tray: Hitachi 2TB 3.5” spinning disk for Audio Files and PT sessions only-
This is the most obvious culprit for PT access problems.
PCIe card - 4 port USB3 interface with 4x4TB Hitachi 3.5“ spinning disk drives connected which have Carbon Copy Clone backups taking place every night, backing up all the other drives. In theory, these drives are never accessed by active apps but they are there spinning all the time.
Just realised this is 11 drives!
Keep in mind that because of all these drives, when I tried to unlock the startup disk dialog to change system drives, I would wait a minute or two for the authorisation dialog to appear UNLESS I did it immediately following the boot up of the Mac Pro. I surmised this was because the drives had no time to spin down. If I worked for a few hours on an audio project and then tried to change I had to wait - probably while all the disks spun up. That put me onto a potential solution for these PT errors.

I also think that if people on MacBook Pros with soldered-in Apple SSDs are having problems, it's because Apple are doing some background tasks on the drive(Trimming?) when the drive thinks its idle(session cache in PT?) Again Avid would have to write a HD poker. But we know that AVID's recommendation is for a separate drive for audio files and that means an external drive for a laptop. Given I am now on Day 5 of seamless PT operation, the more all this makes a kind of sense to me.
I hope this helps
jonh

[lesbrunn]

Originally Posted by The Drummer

Given I am now on Day 5 of seamless PT operation, the more all this makes a kind of sense to me.
I hope this helps
jonh

It does make sense, bro. None of my drives have any automated/scheduled activities, but anything that would help with the 9173 thing would be worth more than its weight in gold.

[treason]

Originally Posted by The Drummer

Let me be clearer:
I surmised this was because the drives had no time to spin down. If I worked for a few hours on an audio project and then tried to change I had to wait - probably while all the disks spun up. That put me onto a potential solution for these PT errors.

Given I am now on Day 5 of seamless PT operation, the more all this makes a kind of sense to me.
I hope this helps
jonh

I have a 2012 MacPro w/2x 512 Apple SSDs (boot and Samples), and 6 WD 6TB Black Caviars on eSATA (connected via PCIe cards) and a FW800 1TB Caviar for DropBox.

I have ALWAYS disabled spin down on all drives (a habit started since the TDM days of PT). For me this is enabled explicitly in my eSATA PCIe card config, and I have also disabled the PUT HARD DRIVES TO SLEEP WHEN POSSIBLE in the Energy Saver Control Panel.

I am the HDX test contributor who passed the OP test - FYI.

[Darryl Ramm]

Originally Posted by treason

I have a 2012 MacPro w/2x 512 Apple SSDs (boot and Samples), and 6 WD 6TB Black Caviars on eSATA (connected via PCIe cards) and a FW800 1TB Caviar for DropBox.



I have ALWAYS disabled spin down on all drives (a habit started since the TDM days of PT). For me this is enabled explicitly in my eSATA PCIe card config, and I have also disabled the PUT HARD DRIVES TO SLEEP WHEN POSSIBLE in the Energy Saver Control Panel.



I am the HDX test contributor who passed the OP test - FYI.

If there is a drive issue here in the system where the drive ping utility seems to fix things it is very likely it is associated with external USB drives (or possibly internal RAID card) in the posters system. External USB (and some RAID setups) may not obey any “do not spin down” settings in the OS. That is exactly/the only thing what the disk utility being used was written to work around. Normal eSATA drives like you have should obey those OS settings.

There undoubtably are/already have been multiple separate issues that cause these CPU errors. It would be great to see if other folks find the disk ping utility helps them.



Sent from my iPhone using Tapatalk

[treason]

Originally Posted by Darryl Ramm

If there is a drive issue here in the system where the drive ping utility seems to fix things it is very likely it is associated with external USB drives (or possibly internal RAID card) in the posters system. External USB (and some RAID setups) may not obey any “do not spin down” settings in the OS. That is exactly/the only thing what the disk utility being used was written to work around. Normal eSATA drives like you have should obey those OS settings.

There undoubtably are/already have been multiple separate issues that cause these CPU errors. It would be great to see if other folks find the disk ping utility helps them.



Sent from my iPhone using Tapatalk

That's great! What my post was meant to convey is the config of a system that appears to be humming along - and confirm that I neither have USB drives, nor a spin down "preference" enabled in my system.

Removing variables from configs is what we're all trying to do here.

[[email protected]]

I wanted to share these 2 screenshots. They're the EXACT same mix template.
One has 52/128 voices active and it's running the CPU @77% (with less errors, mind you)
The other, 68/128 voices, running at 49% with WAY more freezes and errors.
These are THE EXACT same mix templates with the only difference being the track count.

[georges Gholam]

Hi, what’s your mac model, and how ram do you have? And wt’s the size of your HDD? I would suggest checking the buffer size.


Sent from my iPhone using Tapatalk

[[email protected]]

Originally Posted by georges Gholam

Hi, what’s your mac model, and how ram do you have? And wt’s the size of your HDD? I would suggest checking the buffer size.


Sent from my iPhone using Tapatalk

Buffer siZe is the same. 1024. You're missing the point of the post. BOTH these sessions are running off of the same drive with same layout with very different results.
Just to entertain- Mac is a late 2012 12 core MacPro with 24GB of ram.
Record drive is a 1TB SSD with 200GB free.