Unexplainable Number of Drive Reads


Recommended Posts

After attempting to find the cause of this on my own, I believe another set or sets of eyes (or ideas for that matter) may be what the doctor ordered.

I have one drive in my array that has a great deal of reads.  I cleared the statistics just 24 hours ago, and on low usage unRAID box, this drive (disk 20) has managed to climb up to 13 million reads, despite all the other drives being under 400,000 reads (besides disk 2 which contains this weeks programming so that makes sense).

 

I used File Activity app/plugin where this was the result over the same 24 hour span from the time I cleared statistics, was just two files opened.  Hardly explains the 13 million reads I would think?

The logs show nothing of significance, just business as usual.  Ever entry concerning sdh (disk 20) appeared normal and consistent with the other drives in the array.

I'm at a loss.  It's not part of a spin up group (the act of spinning up doesnt increase reads though does it?).    As a result the drive is powered up, having a power on hours count of 3 years now, where disk 18 and 19 were put into the system at the same time, but have power on hours of only 1 year 8 months.  I read that to mean, unnecessary wear on disk 20, but from what?  I dunno.

Attachment showing the drive stats for a 24 hour period (give or take two hours) after clearing stats.


 

 

unexplained_drive_reads.jpg

Edited by grandprix
Link to comment

Another 16 million reads, now up to 29 million reads.  According to File Activity plugin/app, still only the two files that were "open"ed as before the reads climbed up another 16 million.

Driving me a bit bonkers.  That seems like an awful lot of reads, for no obvious reason.  Including the diagnostics, as mentioned my eyes didn't catch anything, but that might be part of the issue as well.  Hoping someone may be able to lead me to some idea why disk 20 is getting hammered with reads.  TIA!

 

tower-diagnostics-20180111-1135.zip

Link to comment
1 hour ago, Squid said:

You can always try starting in safe mode.  Most obvious plugin though that could be responsible is cache_dirs.  Beyond that, disable docker apps one at a time to narrow things down.


Will do.  Cleared the statistics and stopped cache_dirs.   Activity from Sonarr, Radarr, SAB appeared to show up in File Activity logs, as the processes (open, etc) coincided with activity of those docker's logs.

I wonder why if it is cache_dirs why it would hammer up just one disk though?  I admit, I'm ignorant to precisely what cache_dirs does (aside from keeping directory structure in memory?).

Link to comment

No such luck.  Stopped all dockers for 24 hours (actually a little longer, as I simply forgot about them until family went to watch Plex).  Disk 20 still sees an unusually high number of reads.  File Activity plugin for what it's worth, seems to catch all files opened (touched?) fairly well.  I'm no closer to figuring this out unfortunately.  Oddly, with the dockers stopped, not only did it have a high number of reads, but, 42 million reads (mind you, I cleared statistics after stopping all dockers and disabling cache_dirs).  (shrug)

Link to comment
  • 3 weeks later...

Issue still persists, I'm at a total loss.  162 million reads on this drive at the moment, with 42 million reads on the drive that is currently storing the newest media (so is the most popular drive at the moment so to speak).  I'm afraid at this rate, I'm afraid the platters will get a hole worn in them <grin>

 

Link to comment
  • 7 months later...

Hate to 9 month necro this post, but, it seems if you are using a Marvell 88SE9215 based controller (in my case specifically the SYBA/IOCrest PEX40064) that is going bad, it will just do random reads, not quite sure the process but evidently a communication between the card and whatever drive.  CH1 that is.  It crapped out entirely on me two days ago.  Got the replacement today, rebuilding the drive now (after I confirmed via SMART it was fine on another computer using the SATA cable I was using to ensure it wasn't that either).  It's on 24/7 and had been, oh, I guess around the time it began it's multiple reads on CH1 of the card, it was 2 years old?  It was a small PITA, but, I got the stock heatsink off the new one, attached a copper heatsink with long "fins" and attached a fan to it.  Fortunately, it is the last PCIe card in the line up, so the heatsink and fan do not obstruct nor create an obstruction.  I'm hoping to get at least two years out of this one.

Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.