1 hour ago, bally12345 said:

100% something not quite right, tried a quick check in Sabnzbd which all download write to cache and it seem to bottleneck around 7MB/s compared to the 45MB/s I normally get also the pystone score seems very low

 

System performance (Pystone)5911  Intel(R) Xeon(R) CPU E564…
Download folder speed11.5 MB/s  (/data/incomplete)
Complete folder speed8 MB/s  (/data/completed)

 

server-diagnostics-20200113-2331.zip 151.08 kB · 0 downloads

One possibility is that it appears that your downloads share (hard to tell exactly because of anonymizing) is set to use cache: yes (every moves from the cache drive to the array whenever mover runs), and files currently exist on all the drives except for disk 10.  You would get far better performance from that share by relegating it exclusively to the cache drive (or use the array only for overflow [use cache: prefer])

 

Also, while you have dynamix ssd trim installed, in the period from Jan 1 - Jan 13, a trim never ran.  Doesn't mean much by itself, but did you set a schedule for trim to run?

One possibility is that it appears that your downloads share (hard to tell exactly because of anonymizing) is set to use cache: yes (every moves from the cache drive to the array whenever mover runs), and files currently exist on all the drives except for disk 10.  You would get far better performance from that share by relegating it exclusively to the cache drive (or use the array only for overflow [use cache: prefer])  

Also, while you have dynamix ssd trim installed, in the period from Jan 1 - Jan 13, a trim never ran.  Doesn't mean much by itself, but did you set a schedule for trim to run?

 

Pretty sure it always been set to Yes and it just moves it to array when the SSD reaches 70% that was at least the case with the 240GB SSD. 

Looked at SSD Trim schedule and is set to run daily at 04:00 so hopefully it did run.

 

Server been up and down due trying upgrade cpus which didn't happen

 

If I set to prefer won't all downloads just stay on the cache drive?

 

Tried but no faster screenshots added

 

Sent from my SM-G973F using Tapatalk

 

 

 

 

On 1/13/2020 at 10:23 AM, Hoopster said:

Fortunately, the "error" is not usually indicative of an actual problem with the drive.  Many of us have used MX500 SSDs for years and simply disabled reporting of this attribute so we don't get nagged by it.  Of course, that means the SSD could at some point develop a bunch of pending sectors and be failing and you would never know.

Are you disabling the attribute labeled "Attribute=197 Current pending sector count" ?

2 minutes ago, mikesp18 said:

Are you disabling the attribute labeled "Attribute=197 Current pending sector count" ?

Yes

I've just got 1 of these drives and can confirm it's still an issue. I'm mounting via the unassigned devices plugin as well.

Hey guys, 

i just bought 2 of these "Crucial MX500 250GB CT250MX500SSD1(Z)" for a raid1 cache, because they are in-expensive in Germany (~35 Euro a piece) and have at least a bit of a DRAM Cache, sadly within 2 days i got the first error ... do any of you have any recommendation for an alternative while i still can return both of them?

 

(My scenario, at least raid1 of 2 identical ssds, about 250GB in size, should have DRAM Cache; Usage: a bit of copy-cache, a few dockers 5-15 and maybe a vm in the future with gpu-passthrough, just set and forget as a home-lab, the array has 16TB of space with 1 parity disk)

 

Following the example of @allanp81 - Its 12-2020 and still happening, did no firmware upgrade.

Edited December 30, 20205 yr by jammsen

Just disable pending sector monitoring, they are still good devices.

23 minutes ago, JorgeB said:

Just disable pending sector monitoring, they are still good devices.

Is this warning from an old harddisk era or is there a reason why its used on ssds?

If i turn that off, can it come to disk errors and me loosing data unknowingly?

Edited December 30, 20205 yr by jammsen

20 minutes ago, jammsen said:

If i turn that off,

Turn it odd just for those SSDs, it's useless for them.

I know this thread started years ago and a firmware update wasn't originally available.  I did notice on the Crucial site that there is now a firmware update available but it is unclear if it fixes the issue at hand.  Has anyone has tried the new firmware available to see it it resolves the problem?  Both of my MX500 drives are currently on M3CR023 and I've been experiencing this same issue.  I just ran across the firmware update and haven't had a chance to try it yet, so wanted to inquire.  Thanks!!

 

https://www.crucial.com/support/ssd-support/mx500-support

 

 

Edited March 21, 20215 yr by dereitz

2 hours ago, dereitz said:

it it resolves the problem?

It's not mencioned in the fixes, but please let us know.

On 3/21/2021 at 11:25 AM, JorgeB said:

It's not mencioned in the fixes, but please let us know.

Unfortunately, I just noticed this little tidbit on the Crucial website that I hadn't noticed earlier:

The M3CR033 firmware is only compatible with drives that shipped with M3CR032.

Given that, I guess I'm stuck on M3CR023.

Unaid 6.9.0/6.9.1 seems to break turning OFF reporting of pending sectors.  Check-box came back "ON" after update and it won't save in a turned off state.

6 minutes ago, Espressomatic said:

Unaid 6.9.0/6.9.1 seems to break turning OFF reporting of pending sectors.  Check-box came back "ON" after update and it won't save in a turned off state.

There's a bug with SMART settings, it will be fixed next release, for now and if you want you can edit the file manually:

 

 

 

On 3/31/2021 at 10:14 AM, JorgeB said:

There's a bug with SMART settings, it will be fixed next release, for now and if you want you can edit the file manually:

 

 

Unfortunately, still happening with 6.9.2.  What's happening is that any time a disk's settings are changed in the UI, the config file is overwritten with only the settings for that single disk. Editing manually as you suggested should work - until you do anything in the UI elsewhere to overwrite it.

Edited April 12, 20215 yr by Espressomatic

10 hours ago, Espressomatic said:

until you do anything in the UI elsewhere to overwrite it.

That file only contains individual device settings, you can change anything else in the GUI.

11 hours ago, JorgeB said:

That file only contains individual device settings, you can change anything else in the GUI.

 

Maybe I wasn't clear.  I have 2 cache drives.  If I change the SMART monitoring on one of them, that gets written to the configuration file. If I go and change the settings on the second drive, the configuration file is not appended to, it is overwritten with only the settings for this new drive.

 

Using the web UI, it's not currently possible to get configuration data into that file for more than a single drive - on my system. Because of this, the system will continue to complain about pending sectors. Only manually editing the configuration to include both drives prevents the errors.

 

Edited April 13, 20215 yr by Espressomatic

10 minutes ago, Espressomatic said:

If I change the SMART monitoring on one of them, that gets written to the configuration file. If I go and change the settings on the second drive, the configuration file is not appended to, it is overwritten with only the settings for this new drive.

Yes. that why I linked how to do it manually, that's what I did for all my drives, once it's done you don't need to touch it anymore, or do you keep changing the settings?

1 minute ago, JorgeB said:

Yes. that why I linked how to do it manually, that's what I did for all my drives, once it's done you don't need to touch it anymore, or do you keep changing the settings?

 

Settings are sticking after I edited manually, thanks again.  I was only updating the thread to mention that 6.9.2, which was supposed to fix the issue of web UI settings not saving, didn't appear to fix this for me, and I described exactly what's happening to the contents of the file.