BUILD Verification and discussion about new FreeNAS-System

[joeschmuck]

Now I've changed DIMMB1 with DIMMA1 and let the memtest run another few days.

Glad you relocated the DIMM, I hope the failure follows the DIMM so you will have basically an easy RMA experience. It is just much easier than a motherboard replacement.

-Mark

 

[Globalhawk]

Okay, Memtest was running for another 164 hours and there was no error anymore. I think I will continue the burn-in testing process with installing FreeNAS und running some SMART-tests. The HDDs were not connected through the memtest to avoid nonsensical running-hours.

 

[joeschmuck]

It's possible it was a poor DIMM connection, glad things are looking up.

 

[Globalhawk]

Yep, hopefully. No more errors for now.

So here are my next writeups:

• You don't need a frontpanel-cable if you use SuperMicro-boards with SuperMicro chassis. But the cable brought me the understanding of in which direction I had to connect the 16-pin-cable
• I've determined my SAS-ID through the CTRL+C-Panel during boot-procedure (if you are seeing LSI..). If you are in the menu für your LSI-Controller just hit enter and you see the information about your controller.
• Flashing IT-Mode with DOS does not work for this board. You have to use UEFI-Shell (press F11 during POST). Then mount your USB-Drive (or virtual one); mount blk7: (for me); blk7: ; SMC2308T.NSH; Entering the 9-digit-SAS-No. - done
• And here comes another advise: Don't flash your board with the newest firmware (v19). FreeNAS does not support (in the case of any problem) v19. Or with other words, this driver does not match. You have to use v16, which is supported. Otherwise you will see an alert directly on login:
  
  

• Before i've installed FreeNAS, I checked all HDDs with WDIDLE3 through FreeDOS (google for rufus, great tool - I use this also for creating usb-boot-disks for VMWare vSphere 6.0 and so on). Before you boot into FreeDOS you have to put all your HDDs off from your LSI-Controller. You can do this by connecting your HDDs onto the Intel-S-ATA-Ports. After this step you have to change into IDE-Mode. Now you are ready for WDIDLE3 /r in the FreeDOS-CLI. My drives were already at 300 (5 Min) seconds. Everything fine.
• Started smartctl short, conveyance and long test. All tests went OK.
• Today I've created a SSH-Certificate that wasn't as easy as elsewhere. I had to adjust the chmod of $HOME and .SSH but the tricky one was to know the syntax for the public key for my user inside of the FreeNAS-GUI.
• Currently I'm running jgreco's script and it seems to be okay and signs me to go to another work or asleep (due to the estimated time):
  
  

 

[Globalhawk]

Now i've logged in and the test was running for 20 secons until it's completion. That's the first results:

And the results:

Code:

Performing initial serial array read (baseline speeds)
Mon Jul  6 22:48:00 CEST 2015
Mon Jul  6 23:01:30 CEST 2015
Completed: initial serial array read (baseline speeds)

Array's average speed is 149.003 MB/sec per disk

Disk    Disk Size  MB/sec %ofAvg
------- ---------- ------ ------
da0      3815447MB    146     98
da1      3815447MB    147     98
da2      3815447MB    153    103
da3      3815447MB    148     99
da4      3815447MB    152    102
da5      3815447MB    149    100

Performing initial parallel array read
Mon Jul  6 23:01:30 CEST 2015
The disk da0 appears to be 3815447 MB.
Disk is reading at about 146 MB/sec
This suggests that this pass may take around 435 minutes

                   Serial Parall % of
Disk    Disk Size  MB/sec MB/sec Serial
------- ---------- ------ ------ ------
da0      3815447MB    146    146    100
da1      3815447MB    147    147    100
da2      3815447MB    153    154    100
da3      3815447MB    148    148    100
da4      3815447MB    152    152    100
da5      3815447MB    149    150    100

Awaiting completion: initial parallel array read
Tue Jul  7 08:53:31 CEST 2015
Completed: initial parallel array read

Disk's average time is 34314 seconds per disk

Disk    Bytes Transferred Seconds %ofAvg
------- ----------------- ------- ------
da0         4000787030016   35521    104
da1         4000787030016   34803    101
da2         4000787030016   33283     97
da3         4000787030016   34602    101
da4         4000787030016   33646     98
da5         4000787030016   34031     99

Performing initial parallel seek-stress array read
Tue Jul  7 08:53:31 CEST 2015
The disk da0 appears to be 3815447 MB.
Disk is reading at about 140 MB/sec
This suggests that this pass may take around 453 minutes

                   Serial Parall % of
Disk    Disk Size  MB/sec MB/sec Serial
------- ---------- ------ ------ ------
da0      3815447MB    146    161    111
da1      3815447MB    147    140     96
da2      3815447MB    153    171    112
da3      3815447MB    148    167    113
da4      3815447MB    152    205    136
da5      3815447MB    149    163    109

Awaiting completion: initial parallel seek-stress array read
Wed Jul  8 23:17:05 CEST 2015
Completed: initial parallel seek-stress array read

Disk's average time is 106702 seconds per disk

Disk    Bytes Transferred Seconds %ofAvg
------- ----------------- ------- ------
da0         4000787030016  105144     99
da1         4000787030016   97791     92 ++FAST++
da2         4000787030016   75279     71 ++FAST++
da3         4000787030016  130532    122 --SLOW--
da4         4000787030016  117650    110 --SLOW--
da5         4000787030016  113814    107

 

[Globalhawk]

Now i've started the next phase of tests as described by qwertymodo: