Making a first build

[cyberjock]

do NOT mix modular power supply cables. You'll break stuff in the permanent kind of way. To boot, once you charbroil boards like that it's obvious what you did and you'll never get an RMA for your life.

Use only the cables the PSU came with and buy splitters. If yuo need more cables contact the manufacturer. I know Corsair sent me a bunch of free ones a few years ago when I bought a 1000w PSU and needed power for 24 drives. ;)

 

[Matdif]

do NOT mix modular power supply cables. You'll break stuff in the permanent kind of way. To boot, once you charbroil boards like that it's obvious what you did and you'll never get an RMA for your life.

Use only the cables the PSU came with and buy splitters. If yuo need more cables contact the manufacturer. I know Corsair sent me a bunch of free ones a few years ago when I bought a 1000w PSU and needed power for 24 drives. ;)

Ya thanks saw splitters on google so ill Just get a couple of those.

Freenas keeps track of drive serial numbers yes? So I should probably mark each of these so its visible when you open the case?

 

[cyberjock]

Yes. Marking them with serial numbers is a very good idea. Even if it's just the last 4 digits or something, it'll be unique enough to be identified.

 

[Z300M]

hmmm having trouble with the power supply. Can you buy different cords for a modular power supply? Ive never actually had to do this but the stuff that came with this 450w isnt going to cover 8hd's and 2 ssd's.

I suppose I can go to the store tommorow there should be some kind of splitter I would think.

Modular cords are not necessarily interchangeable from one brand to another, even if they have the same number of pins (and not all do: my Antec modular has 6-pin cables, the Rosewill has 8-pin cables): I think I read here about someone that used cables from a different brand and fried something. As undesirable as splitters may be from one viewpoint, they may be safest in your case now.

 

[9C1 Newbee]

I'd modify a modular cord by splicing in some connectors from an old junk power supply. Or cut the connectors off the splitter it's self. If you don't feel comfortable doing that, just use the splitter as intended. I have used many splitters. Make sure they are plugged in nice and tight.

I have gotten a hold of ill fitting connectors. A well placed zip tie can work wonders. I have a few of these examples under the hood of my car where tabs have broken off. A computer probably won't be pulling 1+ g's either.

 

[Ericloewe]

Modular cords are not necessarily interchangeable from one brand to another, even if they have the same number of pins (and not all do: my Antec modular has 6-pin cables, the Rosewill has 8-pin cables): I think I read here about someone that used cables from a different brand and fried something. As undesirable as splitters may be from one viewpoint, they may be safest in your case now.

He didn't, miraculously. Guess the PSU's protections did their job.
In my experience, most brands can't even keep cables consistent within model lines. It would be nice for there to be a standard.

 

[9C1 Newbee]

He didn't, miraculously. Guess the PSU's protections did their job.
In my experience, most brands can't even keep cables consistent within model lines. It would be nice for there to be a standard.

Don't tell me that. I just threw all my Corsair cables in one heap. I even took a few from the other units to power some of the FreeNAS drives. Hope it doesn't bite me in the ass later. I guess I need to research this further.

 

[Ericloewe]

Don't tell me that. I just threw all my Corsair cables in one heap. I even took a few from the other units to power some of the FreeNAS drives. Hope it doesn't bite me in the ass later. I guess I need to research this further.

Corsair CX-M and CS PSUs probably have identical cables, as they're the same CWT platform.
AX650/750/850 have identical cables. Same goes for AX760/860.

However, if everything's working, it'll keep working. If it was going to cause trouble, you'd know that by know (assuming you've turned it on).

 

[9C1 Newbee]

It has been on and working great (knock knock knock). It didn't let the smoke out. I just didn't even consider this variable and just assumed it was all ok. I have two AX760's. I also have an HX750. That is the one that may or may not be an issue.

EDIT: I googled a few threads. Seems the AX and the HX series is NOT interchangeable. Looks like I was damn lucky.

 

[Ericloewe]

The 750W+ Corsair HX PSUs are all CWT, so I doubt their cables are compatible. Good news is that they might not even fit or are, at least, easy to distinguish from the AX cables.

 

[9C1 Newbee]

Oh that is good news. I will have to dig them out and see whats what. How stupid not to have a set standard across an entire brand. Just setting people up to fail.

 

[Ericloewe]

I am the bearer of good news today: Corsair seems to be doing just that with their newer models:

http://www.corsair.com/en/professional-individually-sleeved-dc-cable-kit-type-3-generation-2-black

Corsair PSU Compatibility:

• AXI Series – 1500/1200/860/760
• AX Series – 860/760
• HX Series – 1050/850/750/650
• HXi Series – 1000/850/750
• TXM Series – 850/750/650/550
• CXM Series – 750/600/500/430
• RM Series – 1000/850/750/650/550/450

This implies that all these use the same cables.

An odd omission are the CS PSUs, which should be very similar to the CXMs.

 

[9C1 Newbee]

I don't know what the hell to believe. I have to agree with you that does imply the cords are compatible. But then there is this:

http://forum.corsair.com/v3/showthread.php?t=94322

Cliff Notes: A forum admin on Corsair's own web site, said "Absolutely not". When asked about HX and AX cable compatibility. But the guy didn't have a dog icon so you have to take what he says with a grain of salt.

Another thing I noticed, the kit you have there says * ATX 24PIN CABLE SOLD SEPARATELY *. Maybe all of the cables except the mother board cables are the same between models.

I don't have time to dig my cords out right now. But I will compare them and post what I find. Kind of a one off type problem. But if it saves one motherboard, it is worth noting it.

Sorry for the hijack

 

[cyberjock]

I can tell you that mixing the HX and AX cables is VERY bad. At least, the ones I saw from about 3 years ago. I saw a computer go up in smoke that way. It was very loud, very fast, very smoky, and got some flames. PSU was fine, but nothing on the system was salvageable.

If you look at the cables it says that they physically fit but are NOT compatible. You can take a multimeter and verify that the pins don't match from one end to the other.. hehe. The cables were clearly marked though. It said on the plugs something like "HX1000 only" and such. That's how we realized what my friend had screwed up.

That's why I'm very forceful when I say not to mix cables. If you *know* better, that's one thing. But if you don't the consequences can be... disastrous.

 

[Ericloewe]

I can tell you that mixing the HX and AX cables is VERY bad. At least, the ones I saw from about 3 years ago. I saw a computer go up in smoke that way. It was very loud, very fast, very smoke, and got some flames. PSU was fine, but nothing on the system was salvageable.

If you look at the cables it says that they physically fit but are NOT compatible. You can take a multimeter and verify that the pins don't match from one end to the other.. hehe. The cables were clearly marked though. It said on the plugs something like "HX1000 only" and such. That's how we realized what my friend had screwed up.

That's why I'm very forceful when I say not to mix cables. If you *know* better, that's one thing. But if you don't the consequences can be... disastrous.

Yeah, my AX850 cables all say "AX750/850 only", but they have messed with most of their line-up since then. I did notice that the AX760 cables didn't say something like that, when I built a system with one the other day.

In any case "DO NOT unless you're absolutely sure" is the tl;dr of this conversation. The mere thought of ruining a motherboard (and who knows what else) is scary enough.

 

[9C1 Newbee]

It was very loud, very fast, very smoky, and got some flames.

OMG LMFAO!

 

[cyberjock]

Yeah. When he's being a jerk I always say "so.. how did your last server build go huh?" I get the expected "f*ck you" and life goes on. LOL!

It's a very sore subject because it cost him over $1500 in hardware. Every piece was badly damaged and both visually as well as fragrantly obvious that someone had abused the hardware.

Needless to say, he doesn't mix up those cables anymore. ;)

 

[Matdif]

Ok its hooked up freenas is running and I got a command window open for the root.

Im trying to use https://forums.freenas.org/index.php?threads/how-to-hard-drive-burn-in-testing.21451/ to run tests on the drives.

the first command is giving me /dev/adaX: Unable to detect device type
Please specify device type with the -d option.

Is there a more detailed set of instruction for this somewhere?
Do I have to form the zpool and stuff before I test the drives?

 

[gpsguy]

You need to substitute your device id for adaX. Usually SATA connected drives are ada0, ada1, ...

the first command is giving me /dev/adaX: Unable to detect device type
Please specify device type with the -d option.

 

[Matdif]

Thanks that worked. Why would the Conveyance test return an error.

Oct 26 12:58:41 freenas manage.py: [reporting.rrd:139] Failed to generate graph: ERROR: opening '/var/db/collectd/rrd/localhost/aggregation-cpu-sum/cpu-idle.rrd': No such file or directory
Oct 26 13:01:12 freenas generate_smb4_conf.py: [common.pipesubr:58] Popen()ing: /sbin/sysctl -n 'kern.maxfilesperproc'
Oct 26 13:01:12 freenas generate_smb4_conf.py: [common.pipesubr:58] Popen()ing: /usr/local/bin/pdbedit -d 0 -i smbpasswd:/tmp/tmpEfhrD7 -s /usr/local/etc/smb4.conf -e tdbsam:/var/etc/private/passdb.tdb
Oct 26 13:01:12 freenas generate_smb4_conf.py: [common.pipesubr:58] Popen()ing: /usr/local/bin/net groupmap list
Oct 26 13:01:13 freenas notifier: Performing sanity check on Samba configuration: OK
Oct 26 13:03:42 freenas manage.py: [reporting.rrd:139] Failed to generate graph: ERROR: opening '/var/db/collectd/rrd/localhost/aggregation-cpu-sum/cpu-idle.rrd': No such file or directory
Oct 26 13:08:42 freenas manage.py: [reporting.rrd:139] Failed to generate graph: ERROR: opening '/var/db/collectd/rrd/localhost/aggregation-cpu-sum/cpu-idle.rrd': No such file or directory
Oct 26 13:13:42 freenas manage.py: [reporting.rrd:139] Failed to generate graph: ERROR: opening '/var/db/collectd/rrd/localhost/aggregation-cpu-sum/cpu-idle.rrd': No such file or directory
Oct 26 13:18:42 freenas manage.py: [reporting.rrd:139] Failed to generate graph: ERROR: opening '/var/db/collectd/rrd/localhost/aggregation-cpu-sum/cpu-idle.rrd': No such file or directory
Oct 26 13:23:42 freenas manage.py: [reporting.rrd:139] Failed to generate graph: ERROR: opening '/var/db/collectd/rrd/localhost/aggregation-cpu-sum/cpu-idle.rrd': No such file or directory
Oct 26 14:27:01 freenas manage.py: [reporting.rrd:139] Failed to generate graph: ERROR: opening '/var/db/collectd/rrd/localhost/aggregation-cpu-sum/cpu-idle.rrd': No such file or directory
Oct 26 14:32:01 freenas manage.py: [reporting.rrd:139] Failed to generate graph: ERROR: opening '/var/db/collectd/rrd/localhost/aggregation-cpu-sum/cpu-idle.rrd': No such file or directory
Oct 26 14:37:01 freenas manage.py: [reporting.rrd:139] Failed to generate graph: ERROR: opening '/var/db/collectd/rrd/localhost/aggregation-cpu-sum/cpu-idle.rrd': No such file or directory
Oct 26 14:42:01 freenas manage.py: [reporting.rrd:139] Failed to generate graph: ERROR: opening '/var/db/collectd/rrd/localhost/aggregation-cpu-sum/cpu-idle.rrd': No such file or directory
Oct 26 14:47:01 freenas manage.py: [reporting.rrd:139] Failed to generate graph: ERROR: opening '/var/db/collectd/rrd/localhost/aggregation-cpu-sum/cpu-idle.rrd': No such file or directory
Oct 26 14:52:01 freenas manage.py: [reporting.rrd:139] Failed to generate graph: ERROR: opening '/var/db/collectd/rrd/localhost/aggregation-cpu-sum/cpu-idle.rrd': No such file or directory
Oct 26 14:53:46 freenas generate_smb4_conf.py: [common.pipesubr:58] Popen()ing: /sbin/sysctl -n 'kern.maxfilesperproc'
Oct 26 14:53:46 freenas generate_smb4_conf.py: [common.pipesubr:58] Popen()ing: /usr/local/bin/pdbedit -d 0 -i smbpasswd:/tmp/tmp402FWD -s /usr/local/etc/smb4.conf -e tdbsam:/var/etc/private/passdb.tdb
Oct 26 14:53:46 freenas generate_smb4_conf.py: [common.pipesubr:58] Popen()ing: /usr/local/bin/net groupmap list
Oct 26 14:53:46 freenas notifier: Performing sanity check on Samba configuration: OK
Oct 26 14:57:01 freenas manage.py: [reporting.rrd:139] Failed to generate graph: ERROR: opening '/var/db/collectd/rrd/localhost/aggregation-cpu-sum/cpu-idle.rrd': No such file or directory
Oct 26 15:00:20 freenas manage.py: [reporting.rrd:139] Failed to generate graph: ERROR: opening '/var/db/collectd/rrd/localhost/aggregation-cpu-sum/cpu-idle.rrd': No such file or directory
Oct 26 15:00:23 freenas manage.py: [reporting.rrd:139] Failed to generate graph: ERROR: opening '/var/db/collectd/rrd/localhost/aggregation-cpu-sum/cpu-idle.rrd': No such file or directory
Oct 26 15:05:23 freenas manage.py: [reporting.rrd:139] Failed to generate graph: ERROR: opening '/var/db/collectd/rrd/localhost/aggregation-cpu-sum/cpu-idle.rrd': No such file or directory

Thats from my log and on the terminal no matter what drive I try (out of 9) they all throw the same error to the conveyance test.

CAM status: ATA Status Error
ATA stats: (DRDY SERV ERR), error: 04 (ABRT )

then failed: Input/output error

That repeating error about the file not being able to be created has been going since freenas came up and UI figured it would be corrected when I set things up. No idea why all the drives wouldn't respond to the conveyance test unless I need to mount them somehow first. They all took the short smart test without trouble.