Any way to test individual hard drive speed?

Status
Not open for further replies.

RichTJ99

Patron
Joined
Sep 12, 2013
Messages
384
Hi,

I have a number of drives & feel like one specific drive (not sure which one) is slowing the freenas box down. Of course this is just a gut feeling. Is there a way to test specific drive health, RPM, etc?

Would mixing 5400 / 7200 RPM drives together cause a noticeable slowdown (say if the 5400 was spinning at 4000?).


Thanks,
Rich
 

jgreco

Resident Grinch
Joined
May 29, 2011
Messages
18,680
The 5400 one won't be spinning at 4000 unless it's nonfunc.

Check around for solnet-array-test v2 which will report drive speeds in a useful manner.
 

RichTJ99

Patron
Joined
Sep 12, 2013
Messages
384
Thanks, with some reading it looks like:

Code:

wget ftp://ftp.sol.net/incoming/solnet-array-test-v2.sh
ls -la
chmod +x solnet-array-test-v2.sh
./solnet-array-test-v2.sh


My testing is underway, not exactly sure how long it takes though.

Thanks,
Rich
 

RichTJ99

Patron
Joined
Sep 12, 2013
Messages
384
So something happened last night during the test - the server was stuck in a boot loop. A full power down seemed to fix it.

Before the crash i have this:

Code:
Selected disks: ada0 ada1 ada2 ada3 ada4 ada5
<WDC WD20EARX-00PASB0 51.0AB51>    at scbus1 target 0 lun 0 (ada0,pass0)
<WDC WD20EZRX-00D8PB0 80.00A80>    at scbus2 target 0 lun 0 (ada1,pass1)
<SAMSUNG HD203WI 1AN10003>         at scbus3 target 0 lun 0 (ada2,pass2)
<Hitachi HDS722020ALA330 JKAOA28A>  at scbus4 target 0 lun 0 (ada3,pass3)
<SAMSUNG HD204UI 1AQ10001>         at scbus5 target 0 lun 0 (ada4,pass4)
<Hitachi HDS722020ALA330 JKAOA3MA>  at scbus6 target 0 lun 0 (ada5,pass5)
Is this correct? (y/N): y
Performing initial serial array read (baseline speeds)
Fri Jan  8 17:41:04 EST 2016
Fri Jan  8 17:54:34 EST 2016
Completed: initial serial array read (baseline speeds)

Array's average speed is 114.22 MB/sec per disk

Disk    Disk Size  MB/sec %ofAvg
------- ---------- ------ ------
ada0     1907729MB     99     87 --SLOW--
ada1     1907729MB    142    124 ++FAST++
ada2     1907729MB    100     88 --SLOW--
ada3     1907729MB    115    101
ada4     1907729MB    115    101
ada5     1907729MB    114    100

Performing initial parallel array read
Fri Jan  8 17:54:34 EST 2016
The disk ada0 appears to be 1907729 MB.
Disk is reading at about 99 MB/sec
This suggests that this pass may take around 321 minutes

                   Serial Parall % of
Disk    Disk Size  MB/sec MB/sec Serial
------- ---------- ------ ------ ------
ada0     1907729MB     99    100    101
ada1     1907729MB    142    141     99
ada2     1907729MB    100    100    100
ada3     1907729MB    115    115    100
ada4     1907729MB    115    100     87 --SLOW--
ada5     1907729MB    114    115    100

Awaiting completion: initial parallel array read
Sat Jan  9 01:00:56 EST 2016
Completed: initial parallel array read

Disk's average time is 21851 seconds per disk

Disk    Bytes Transferred Seconds %ofAvg
------- ----------------- ------- ------
ada0        2000398934016   25582    117 --SLOW--
ada1        2000398934016   17773     81 ++FAST++
ada2        2000398934016   25199    115 --SLOW--
ada3        2000398934016   21035     96
ada4        2000398934016   20225     93
ada5        2000398934016   21289     97

Performing initial parallel seek-stress array read
Sat Jan  9 01:00:56 EST 2016
The disk ada0 appears to be 1907729 MB.
Disk is reading at about 87 MB/sec
This suggests that this pass may take around 367 minutes

                   Serial Parall % of
Disk    Disk Size  MB/sec MB/sec Serial
------- ---------- ------ ------ ------
ada0     1907729MB     99     92     93
ada1     1907729MB    142    134     94
ada2     1907729MB    100     11     11
ada3     1907729MB    115    115    100
ada4     1907729MB    115     18     16
ada5     1907729MB    114    109     96

Awaiting completion: initial parallel seek-stress array read
 

jgreco

Resident Grinch
Joined
May 29, 2011
Messages
18,680
The parallel seek-stress test seems to show that your Spinpoint drives aren't performing well when pressed for multiple simultaneous operations. Maybe NCQ is disabled on those two drives?
 

Mirfster

Doesn't know what he's talking about
Joined
Oct 2, 2015
Messages
3,215
Thanks, with some reading it looks like:
Code:
wget ftp://ftp.sol.net/incoming/solnet-array-test-v2.sh
ls -la
chmod +x solnet-array-test-v2.sh
./solnet-array-test-v2.sh

Thinking that it wouldn't be a bad idea to have these steps in the original thread (https://forums.freenas.org/index.php?threads/building-burn-in-and-testing-your-freenas-system.17750/). Maybe right after the part that says:
To run it, download it onto a FreeNAS box and execute it as root. It will give you a simple menu

Of course, the "ls -la" step could be left out since it is not really needed..
 

RichTJ99

Patron
Joined
Sep 12, 2013
Messages
384
The parallel seek-stress test seems to show that your Spinpoint drives aren't performing well when pressed for multiple simultaneous operations. Maybe NCQ is disabled on those two drives?

How would I check if the NCQ is disabled? Is that something i can see from ssh?
 

jgreco

Resident Grinch
Joined
May 29, 2011
Messages
18,680
Yup, "camcontrol identify <devicename>" and look for NCQ. Reducing the number of tags in NCQ is more or less disabling it, so if there's a really low number, ... bad.
 
Status
Not open for further replies.
Top