So I was running low on space and decided it's time to start swapping in 4tb drives.  Upgraded to 5.0, got some Seagates, precleared them, and swapped in my new parity drive.  It's going fine but I noticed something odd.  The drives connected to my AOC-SASLP-MV8 card have nearly twice the reads of the drives connected to the SATA ports on the Supermicro X7SPA-H motherboard.  Anyone know what's up with that?

 

It is the way the Linux drivers batch the reads and writes to the drives. 

 

This comment was based on a post in another thread that I am unable to find right now and my interpretation of what I read in that post.

Normal -- nothing to worry about.

 

I had a long discussion about this with Tom several months ago.  I can't find the specific thread right now (probably the same thread Bob was trying to find) ... but the bottom line is it's normal behavior and nothing to worry about.

 

Normal -- nothing to worry about.

 

I had a long discussion about this with Tom several months ago.  I can't find the specific thread right now (probably the same thread Bob was trying to find) ... but the bottom line is it's normal behavior and nothing to worry about.

Yes it was.  I was going to look through your posts but you have so many I didn't want to take the time.

Thanks.  "It's not a bug.  It's a feature."    I wasn't worried so much as curious since it seems to be cruising along fine and I've had this setup for years without noticing any problems (or the discrepancy).  This is probably the first time that I've booted and gone directly into a parity build with the 8 port controller attached so the difference in read values wasn't as obvious until now.

Thanks.  "It's not a bug.  It's a feature." 

 

I wouldn't characterize it like that.    It's truly not an issue at all.  Basically, UnRAID is reporting the number of Read "requests" => and due to the way these are buffered, the numbers can be significantly different -- even on the same controller.    I get widely varying numbers on both of my servers ... and on my smaller server all of the disks are on the same chipset-based motherboard SATA ports.

 

I have noticed that on my test systems, the older ones (with IDE drives) have identical read numbers for all drives;  and I THINK that if you set your SATA mode to IDE you'd see the same thing.  [i haven't actually tried that]

 

But as I noted above, it's simply nothing to be concerned about.

 

That's why I put it in quotes.  With a smiley.