Register a SA Forums Account here!
JOINING THE SA FORUMS WILL REMOVE THIS BIG AD, THE ANNOYING UNDERLINED ADS, AND STUPID INTERSTITIAL ADS!!!

You can: log in, read the tech support FAQ, or request your lost password. This dumb message (and those ads) will appear on every screen until you register! Get rid of this crap by registering your own SA Forums Account and joining roughly 150,000 Goons, for the one-time price of $9.95! We charge money because it costs us money per month for bills, and since we don't believe in showing ads to our users, we try to make the money back through forum registrations.
 
  • Post
  • Reply
Lowen SoDium
Jun 5, 2003

Highen Fiber
Clapping Larry

Gozinbulx posted:

How many actual 3.5 drives can you fit in that case. I get confused when the specs say it has x number of bays and then you guys post things like "4x3 5.25 to 3.5 drive bay" which I guess lets you somehow put more drives? I'm sorry just a bit confused.

Also, over in the HTPC thread, someone said they simply use WHS2011 and a utility that pools the disks (not the one supplied with window server) and that it has worked well for them.

I am willing to consider ZFS and unRaid and all that stuff but man, that sounds so much easier. Someone here convince me otherwise.

It has a hotswap cage that holds five 3.5" SATA drives, and there is space in the bottom for one 3.5" or 2.5" drives. But if the bottom spots might conflict with PCI-Express cards that are double wide. It doesn't have any external drive or 5.25" bays.

Adbot
ADBOT LOVES YOU

widds2v
Aug 23, 2004
Sorry, I'm socially retarded.
Anyone with Synology disaster recovery experience?

My dad has a DS411j that was running JBOD (I know, I know) and his Disk 1 failed. It's dead dead, took it out and threw it in a SATA reader.. just clicks for 30 seconds and then dies. Disks 2-4 seem fine as per the Storage Manager but since 1 is down the whole volume is full on Crashed.

My original plan was just to pull the data off the individual drives with a USB SATA reader, but some Googling as revealed very little about using Windows to read extfs drives that were used in a JBOD and mounting them. Then I was going to replace Disk 1 & 4 with 3TB WD Red drives (original Disk 4 was a 2TB drive, Disk 2 & 3 are 3TB Seagate drives) and get him setup with SHR/RAID5. However, with Volume 1 crashed, File Station shows zero Shared Folders or files. I've sent off an email to Synology support but with the weekend and the ~3 business days to reply I'd really like to not waste the next week not doing anything with the server.

Ideas on how I could access the Volume and start pulling files off the three remaining good drives?

Septimus
Aug 30, 2003
Wasabi? Why not!

widds2v posted:

Anyone with Synology disaster recovery experience?

My dad has a DS411j that was running JBOD (I know, I know) and his Disk 1 failed. It's dead dead, took it out and threw it in a SATA reader.. just clicks for 30 seconds and then dies. Disks 2-4 seem fine as per the Storage Manager but since 1 is down the whole volume is full on Crashed.

My original plan was just to pull the data off the individual drives with a USB SATA reader, but some Googling as revealed very little about using Windows to read extfs drives that were used in a JBOD and mounting them. Then I was going to replace Disk 1 & 4 with 3TB WD Red drives (original Disk 4 was a 2TB drive, Disk 2 & 3 are 3TB Seagate drives) and get him setup with SHR/RAID5. However, with Volume 1 crashed, File Station shows zero Shared Folders or files. I've sent off an email to Synology support but with the weekend and the ~3 business days to reply I'd really like to not waste the next week not doing anything with the server.

Ideas on how I could access the Volume and start pulling files off the three remaining good drives?


Try doing the freezer trick on the dead hdd? Maybe you can miraculously get it running and clone it.

edit: after reading about the freezer trick and "stiction" it seems to be a bad idea!

Septimus fucked around with this message at 04:30 on Apr 6, 2013

widds2v
Aug 23, 2004
Sorry, I'm socially retarded.

Septimus posted:

Try doing the freezer trick on the dead hdd? Maybe you can miraculously get it running and clone it.

Well I had it in the freezer for about an hour before putting the drives back in the Synology, as I needed to turn it on to get the DSM version number and stuff for a customer support ticket. The drive was pretty cold when I booted the thing up and it still didn't read it and had the same errors.

widds2v
Aug 23, 2004
Sorry, I'm socially retarded.
Would removing the volume and then remaking it (JBOD) with Disks 2-4 format the disks? I know RAID would format it, but JBOD?

Longinus00
Dec 29, 2005
Ur-Quan

widds2v posted:

Anyone with Synology disaster recovery experience?

My dad has a DS411j that was running JBOD (I know, I know) and his Disk 1 failed. It's dead dead, took it out and threw it in a SATA reader.. just clicks for 30 seconds and then dies. Disks 2-4 seem fine as per the Storage Manager but since 1 is down the whole volume is full on Crashed.

My original plan was just to pull the data off the individual drives with a USB SATA reader, but some Googling as revealed very little about using Windows to read extfs drives that were used in a JBOD and mounting them. Then I was going to replace Disk 1 & 4 with 3TB WD Red drives (original Disk 4 was a 2TB drive, Disk 2 & 3 are 3TB Seagate drives) and get him setup with SHR/RAID5. However, with Volume 1 crashed, File Station shows zero Shared Folders or files. I've sent off an email to Synology support but with the weekend and the ~3 business days to reply I'd really like to not waste the next week not doing anything with the server.

Ideas on how I could access the Volume and start pulling files off the three remaining good drives?

JBOD is not like having files individually on each disk, it's having the entire filesystem spread over the disks. This is a huge difference. Imagine a hard drive where suddenly you couldn't read from one of the platters anymore (yes this is not an accurate analogy but it's close enough).

That being said, it might still be possible to recover the data. If the controller/OS driver that created the JBOD was only concatenating the disks together and not doing anything sufficiently fancy then a tool which scans raw partitions to find ext metadata blocks should be able to recover stuff. A good question might be whether or not you need to be working on a recreated JBOD to have any hope for the addresses to line up. In either case I doubt you can do the necessary data recovery on the synology itself. If you're not comfortable with linux and you really want this data then you might consider paying someone else to do it.

widds2v posted:

Would removing the volume and then remaking it (JBOD) with Disks 2-4 format the disks? I know RAID would format it, but JBOD?

You can create new raid volumes without reformatting, it all depends on the software. You really shouldn't do this on the synology because you have no control over what it's doing.

underlig
Sep 13, 2007

widds2v posted:

I've sent off an email to Synology support but with the weekend and the ~3 business days to reply I'd really like to not waste the next week not doing anything with the server.
I would recommend waiting until Synology support has ssh;d into it to check it out before trying anything on your own.

an actual cat irl
Aug 29, 2004

Wow - HP Microserver rebates are pretty lovely now compared to when I bought my N36, about 18 months ago. I think I paid about £100 after the rebate came in, but new models only have a £50 rebate and still cost £200+ when all is said and done.

I hope HP do a bigger rebate again at some point this year; I could really do with another one.

FISHMANPET
Mar 3, 2007

Sweet 'N Sour
Can't
Melt
Steel Beams
So I'm not sure what to do with my pools:
code:
  pool: storage
 state: DEGRADED
status: One or more devices has experienced an error resulting in data
        corruption.  Applications may be affected.
action: Restore the file in question if possible.  Otherwise restore the
        entire pool from backup.
   see: [url]http://www.sun.com/msg/ZFS-8000-8A[/url]
 scan: resilvered 2.30T in 19h13m with 517 errors on Sun Apr  7 20:30:22 2013
config:

        NAME             STATE     READ WRITE CKSUM
        storage          DEGRADED     0     0 1.00K
          raidz1-0       DEGRADED     0     0 2.01K
            c7t2d0       REMOVED      0     0     0
            c7t1d0       ONLINE       0     0     0
            replacing-2  DEGRADED     0     0     0
              c7t0d0     DEGRADED     0     0     0  too many errors
              c8t4d0     FAULTED      0     0     0  corrupted data
            spare-3      ONLINE       0     0     0
              c7t3d0     ONLINE       0     0     0
              c8t4d0     ONLINE       0     0     0
            c8t3d0       DEGRADED     0     0     0  too many errors
          raidz1-1       ONLINE       0     0     0
            c7t7d0       ONLINE       0     0     0
            c7t4d0       ONLINE       0     0     0
            c7t5d0       ONLINE       0     0     0
            c7t6d0       ONLINE       0     0     0
            c8t5d0       ONLINE       0     0     0
        spares
          c8t4d0         INUSE     currently in use
Not pictured, 517 data errors. They're all either files I don't want or files I can easily replace, so I'm not worried about that. What I want is to replace c8t4d0, it was a disk I borrowed during an RMA and now I have the good disk back. But I'm not sure if the disk is in use in the replacing-2 vdev or spare-3. And I ran a 'zpool clear storage' to clear the errors, and it's causing my disk to resliver. And basically I have no idea what's going on anymore.

Combat Pretzel
Jun 23, 2004

No, seriously... what kurds?!
Doesn't the Intel RST driver support hot swapping? I've enabled the specific port in the BIOS and slapped the Intel driver on the system. Yet I don't have the Safely Remove option. I've disabled the write cache on the port and can pull the drive at will, but I'd rather want to spin it down before pulling it.

It's a P67 based board, running Windows 8 and Intel RST 12.x.

--edit: Well, I've ditched the Intel driver and went back to Microsoft's driver. I get the remove option, but the drive doesn't spin down. I guess I'll keep the MS driver, because there seems evidence that it's slightly faster for SSD drives. I've also tried HotSwap!, the drive doesn't spin down. You can hear it start spinning down, just to go back to full spindle speed again.

Combat Pretzel fucked around with this message at 17:53 on Apr 10, 2013

Gozinbulx
Feb 19, 2004
So I potentially can get a free one of these: http://www.dell.com/us/dfb/p/precision-690/pd

I am mostly interested in the case. It claim to have:

4 internal 3.5" hard disk drive bays
3 external 5.25" optical bays, one of which can accommodate a fifth HDD
1 external 3.5" flex bay for floppy or media card reader

I can put one of these 3x4 converters in the 5.25 and put an hd in the floppy? Coming to a total of 9 drives? Is this correct?

IT Guy
Jan 12, 2010

You people drink like you don't want to live!

Gozinbulx posted:

So I potentially can get a free one of these: http://www.dell.com/us/dfb/p/precision-690/pd

I am mostly interested in the case. It claim to have:

4 internal 3.5" hard disk drive bays
3 external 5.25" optical bays, one of which can accommodate a fifth HDD
1 external 3.5" flex bay for floppy or media card reader

I can put one of these 3x4 converters in the 5.25 and put an hd in the floppy? Coming to a total of 9 drives? Is this correct?

I have these at work. I can confirm tomorrow.

Galler
Jan 28, 2008


Sounds about right from what I remember. If IT Guy doesn't get to it I can confirm on Friday.

IOwnCalculus
Apr 2, 2003





FISHMANPET posted:

So I'm not sure what to do with my pools:

Not pictured, 517 data errors. They're all either files I don't want or files I can easily replace, so I'm not worried about that. What I want is to replace c8t4d0, it was a disk I borrowed during an RMA and now I have the good disk back. But I'm not sure if the disk is in use in the replacing-2 vdev or spare-3. And I ran a 'zpool clear storage' to clear the errors, and it's causing my disk to resliver. And basically I have no idea what's going on anymore.

You have to delete any corrupted files or ZFS will bitch about them until the heat death of the universe. I know that in NAS4Free, when a drive drops out the old entry stops being /dev/da8 or c2t4d8 or whatever nomenclature your OS uses, and starts being some enormous numeric string. After replacing one with a spare, you can remove the old entry:

code:
2013-03-24.21:10:42 zpool add aggregate spare da8
2013-03-30.13:02:49 zpool replace aggregate 13375166882362914288 da8
2013-03-30.13:16:57 zpool detach aggregate 13375166882362914288
The giant number starting with 1337 is the drive formerly at /dev/ada2 that dropped off. I actually need to RMA that thing again, it's still in warranty...

Which leaves the current pool at:

code:
  pool: aggregate
 state: ONLINE
  scan: scrub repaired 0 in 13h40m with 0 errors on Sat Apr  6 14:40:24 2013
config:

	NAME        STATE     READ WRITE CKSUM
	aggregate   ONLINE       0     0     0
	  raidz1-0  ONLINE       0     0     0
	    da4     ONLINE       0     0     0
	    da3     ONLINE       0     0     0
	    da5     ONLINE       0     0     0
	  raidz1-1  ONLINE       0     0     0
	    da1     ONLINE       0     0     0
	    da2     ONLINE       0     0     0
	    ada1    ONLINE       0     0     0
	  raidz1-2  ONLINE       0     0     0
	    da6     ONLINE       0     0     0
	    da7     ONLINE       0     0     0
	    da8     ONLINE       0     0     0
	cache
	  ada0      ONLINE       0     0     0

errors: No known data errors
Only thing is in your situation I don't know how you specify which drive you're detaching - whether ZFS will be smart enough to remove the old entry marked as corrupted, or if it will try to remove the current one (which, in NAS4Free, will error out rather than letting you murder your own array).

thideras
Oct 27, 2010

Fuck you, I'm a tree.
Fun Shoe
Has anyone been running ZFS on Linux (not FreeNAS, etc), specifically CentOS? With the changes to my server(s) setup, I'm seriously considering it for my file server as I have quite the spread of disks. I'm mainly worried about stability and integrity of files. It is either this or mdadm as my 8708EM2 controller is destined for my Dell R710, and the second 8708EM2 that just arrived is going in the Dell R510. It is going to be interesting getting the files swapped over as the main share array is the nine 2 TB drives below. Once I disconnect the 8708EM2 from the SAS expander card in the server, that is one disk too many for the RAID controller. I'll need to do some data cleaning and moving.

Current disks:
3x 36 GB 10k RPM Cheetah (OS drives, RAID 1)
7x 1 TB Hitachi 7k1000
9x 2 TB Hitachi 5k3000
8x 1.5 TB Seagate Barracuda



Also, have some server hardware porn. The faceplates for the R710 and R510 arrived a few days ago.


(Ignore the warning light on the R510, it is just complaining the backplane isn't connected. No crap, server, you don't have a RAID controller, shut up.)



FISHMANPET
Mar 3, 2007

Sweet 'N Sour
Can't
Melt
Steel Beams
Ugh, turns out my problems are even worse.

1) My resliver never finishes, it just keeps restarting. I've seen it get as high as 60% but this morning it restarted around 11%
2) One of my drives keeps going offline. It doesn't show any read/write/checksum errors, it just goes offline, and bringing it online restarts the resliver.

Dmesg has a bunch of SCSI stuff, but I have no idea how to translate it into anything useful. Here's a few entries:

code:
Apr 11 09:02:53 <servername> scsi: [ID 365881 kern.info] /pci@0,0/pci1002,597a@4/pci15d9,a580@0 (mpt0):
Apr 11 09:02:53 <servername>    Log info 0x31080000 received for target 2.
Apr 11 09:02:53 <servername>    scsi_status=0x0, ioc_status=0x804b, scsi_state=0x0
Apr 11 09:02:53 <servername> scsi: [ID 365881 kern.info] /pci@0,0/pci1002,597a@4/pci15d9,a580@0 (mpt0):
Apr 11 09:02:53 <servername>    Log info 0x31080000 received for target 2.
Apr 11 09:02:53 <servername>    scsi_status=0x0, ioc_status=0x804b, scsi_state=0x0
Apr 11 09:02:53 <servername> scsi: [ID 365881 kern.info] /pci@0,0/pci1002,597a@4/pci15d9,a580@0 (mpt0):
Apr 11 09:02:53 <servername>    Log info 0x31080000 received for target 2.
Apr 11 09:02:53 <servername>    scsi_status=0x0, ioc_status=0x804b, scsi_state=0x0
Apr 11 09:02:56 <servername> scsi: [ID 365881 kern.info] /pci@0,0/pci1002,597a@4/pci15d9,a580@0 (mpt0):
Apr 11 09:02:56 <servername>    Log info 0x31080000 received for target 2.
Apr 11 09:02:56 <servername>    scsi_status=0x0, ioc_status=0x804b, scsi_state=0x0
Apr 11 09:02:59 <servername> scsi: [ID 365881 kern.info] /pci@0,0/pci1002,597a@4/pci15d9,a580@0 (mpt0):
Apr 11 09:02:59 <servername>    Log info 0x31080000 received for target 2.
Apr 11 09:02:59 <servername>    scsi_status=0x0, ioc_status=0x804b, scsi_state=0x0
Apr 11 09:03:06 <servername> scsi: [ID 365881 kern.info] /pci@0,0/pci1002,597a@4/pci15d9,a580@0 (mpt0):
Apr 11 09:03:06 <servername>    Log info 0x31080000 received for target 2.
Apr 11 09:03:06 <servername>    scsi_status=0x0, ioc_status=0x804b, scsi_state=0x0
Apr 11 09:03:09 <servername> scsi: [ID 365881 kern.info] /pci@0,0/pci1002,597a@4/pci15d9,a580@0 (mpt0):
Apr 11 09:03:09 <servername>    Log info 0x31080000 received for target 2.
Apr 11 09:03:09 <servername>    scsi_status=0x0, ioc_status=0x804b, scsi_state=0x0
Apr 11 09:03:13 <servername> scsi: [ID 365881 kern.info] /pci@0,0/pci1002,597a@4/pci15d9,a580@0 (mpt0):
Apr 11 09:03:13 <servername>    Log info 0x31080000 received for target 2.
Apr 11 09:03:13 <servername>    scsi_status=0x0, ioc_status=0x804b, scsi_state=0x0
This is all runing on Solaris 11 Express and I upgraded the pools so I'm stuck with the OS.

IT Guy
Jan 12, 2010

You people drink like you don't want to live!

Gozinbulx posted:

So I potentially can get a free one of these: http://www.dell.com/us/dfb/p/precision-690/pd

I am mostly interested in the case. It claim to have:

4 internal 3.5" hard disk drive bays
3 external 5.25" optical bays, one of which can accommodate a fifth HDD
1 external 3.5" flex bay for floppy or media card reader

I can put one of these 3x4 converters in the 5.25 and put an hd in the floppy? Coming to a total of 9 drives? Is this correct?

Hey sorry, we actually don't have these. We run an older version of them with a slightly different case. However, I can only assume the website is correct.

poverty goat
Feb 15, 2004



Is there any sane way to increase bandwidth beyond 1gbps between a computer and a freeNAS box sitting next to it? It's not necessary in any capacity, just for fun at home, so a pair of 10gbps NICs aren't really an option unless there's a cheap way to get them refurbished or something. My spare parts NAS is pretty overkill: Phenom II 955BE, 5x1.5tb RAIDZ2, 120gb Agility 3 as cache because I don't trust it as actual storage anywhere in the house, and ZIL mirrored across a conveniently paired set of 64gb Agility 1s. I am currently saturating the 1gbps link in most situations (except for writes from remote mechanical drives :v: ) and who doesn't hate a ceiling like that? Realistically I should be able to get close to 3gbps reading cached data if I can find a pipe big enough to move all of it. Is there any cost effective way to do this?

FISHMANPET
Mar 3, 2007

Sweet 'N Sour
Can't
Melt
Steel Beams
It's called Link Aggregation, and a quick Google says that FreeNAS supports it. You'll need some kind of managed switch that can do the same protocol you're using on your server. If you want to spread that bandwidth across multiple clients you don't need to do anything special beyond configuring the switch and the server, but if you want that increased bandwidth from point to point you'll need to configure link aggregation on your client as well. Expect to run a lot of cables.

poverty goat
Feb 15, 2004



FISHMANPET posted:

It's called Link Aggregation, and a quick Google says that FreeNAS supports it. You'll need some kind of managed switch that can do the same protocol you're using on your server. If you want to spread that bandwidth across multiple clients you don't need to do anything special beyond configuring the switch and the server, but if you want that increased bandwidth from point to point you'll need to configure link aggregation on your client as well. Expect to run a lot of cables.

Is there any way to do it directly point-to-point crossover style?

FISHMANPET
Mar 3, 2007

Sweet 'N Sour
Can't
Melt
Steel Beams
I suppose it's possible, but I've never thought to research it.

Moey
Oct 22, 2010

I LIKE TO MOVE IT

gggiiimmmppp posted:

Is there any way to do it directly point-to-point crossover style?

Are you really exceeding 1gb bandwidth to a single client? And also only need one client hitting that data?

DrDork
Dec 29, 2003
commanding officer of the Army of Dorkness

gggiiimmmppp posted:

Is there any sane way to increase bandwidth beyond 1gbps between a computer and a freeNAS box sitting next to it?
The short answer to your question is "no." You can use LAG, but that only helps you out in cases where you've got multiple computers connecting to the server--it won't speed up that file transfer you're looking for. Even if you LAG the hell out of everything and have LAG support on your server, switch (if any), and client, it still won't speed up the file transfer: it'll let you run multiple transfers at high speed, but any single one will still be limited to 1Gbps. The rationale for this being (as far as I understand it, anyhow) that the TCP stack gets really upset at out-of-order frames. It can handle a few here and there, but jumping around between multiple physical links simply takes too much work and CPU overhead. Hence why everyone looking for faster point-to-point networking solutions goes for faster physical link speeds, rather than just LAGing a bunch together.

poverty goat
Feb 15, 2004



Moey posted:

Are you really exceeding 1gb bandwidth to a single client? And also only need one client hitting that data?

Am I really exceeding 1gbps? Yes
code:
[root@freenas] ~# dd if=/dev/zero of=/mnt/main/ddfile bs=1024 count=50000
50000+0 records in
50000+0 records out
51200000 bytes transferred in 0.342759 secs (149376035 bytes/sec)
[root@freenas] ~# dd if=/mnt/main/ddfile of=/dev/zero bs=1024 count=50000
50000+0 records in
50000+0 records out
51200000 bytes transferred in 0.137000 secs (373722399 bytes/sec)
(it's all samsung 830s on the receiving end except for externals)
Do I need to? No, not really, but it'd be fun and actually handy occasionally for VM stuff or encoding video or
Only one client? Everything else is bottlenecked by wifi including the xbmc which is pretty much the most demanding thing and it can do 1080p over the wifi just fine so I don't give a gently caress. Right now they're both on the same desk plugged into an e2000 running Tomato.

DrDork posted:

The short answer to your question is "no." You can use LAG, but that only helps you out in cases where you've got multiple computers connecting to the server--it won't speed up that file transfer you're looking for. Even if you LAG the hell out of everything and have LAG support on your server, switch (if any), and client, it still won't speed up the file transfer: it'll let you run multiple transfers at high speed, but any single one will still be limited to 1Gbps. The rationale for this being (as far as I understand it, anyhow) that the TCP stack gets really upset at out-of-order frames. It can handle a few here and there, but jumping around between multiple physical links simply takes too much work and CPU overhead. Hence why everyone looking for faster point-to-point networking solutions goes for faster physical link speeds, rather than just LAGing a bunch together.

I expected that the LAG answer would be something like that.

I guess there's no way to get eSATA back out of it? :v:

poverty goat fucked around with this message at 00:15 on Apr 12, 2013

Longinus00
Dec 29, 2005
Ur-Quan

gggiiimmmppp posted:

Am I really exceeding 1gbps? Yes
code:
[root@freenas] ~# dd if=/dev/zero of=/mnt/main/ddfile bs=1024 count=50000
50000+0 records in
50000+0 records out
51200000 bytes transferred in 0.342759 secs (149376035 bytes/sec)
[root@freenas] ~# dd if=/mnt/main/ddfile of=/dev/zero bs=1024 count=50000
50000+0 records in
50000+0 records out
51200000 bytes transferred in 0.137000 secs (373722399 bytes/sec)
(it's all samsung 830s on the receiving end except for externals)
Do I need to? No, not really, but it'd be fun and actually handy occasionally for VM stuff or encoding video or
Only one client? Everything else is bottlenecked by wifi including the xbmc which is pretty much the most demanding thing and it can do 1080p over the wifi just fine so I don't give a gently caress. Right now they're both on the same desk plugged into an e2000 running Tomato.


I expected that the LAG answer would be something like that.

I guess there's no way to get eSATA back out of it? :v:

It's not turn-key and it won't work with freenas but it's certainly possible.

http://multipath-tcp.org/pmwiki.php?n=Main.50Gbps

thebigcow
Jan 3, 2001

Bully!
If you're feeling crazy you can get mellanox, qlogic, or intel infiniband cards on ebay. The obvious problems are operating system support, finding cables, and finding new cards when one of them gives up on you.

BlankSystemDaemon
Mar 13, 2009



LAGG (link aggregation) just requires multiple NICs, a Cisco SLM2008T at ~$100, and is super easy to setup: http://doc.freenas.org/index.php/Link_Aggregations

I have it running here at home with a dual-port NIC from Intel.

BlankSystemDaemon fucked around with this message at 08:02 on Apr 12, 2013

Galler
Jan 28, 2008


Gozinbulx posted:

So I potentially can get a free one of these: http://www.dell.com/us/dfb/p/precision-690/pd

I am mostly interested in the case. It claim to have:

4 internal 3.5" hard disk drive bays
3 external 5.25" optical bays, one of which can accommodate a fifth HDD
1 external 3.5" flex bay for floppy or media card reader

I can put one of these 3x4 converters in the 5.25 and put an hd in the floppy? Coming to a total of 9 drives? Is this correct?

I checked one of our 690s and your bay counts are correct. However the 5.25 bays have metal dividers inside them so I don't believe you could put in a 3x4 converter without some demeling. I can post some pictures when I get home if you want.

Gozinbulx
Feb 19, 2004

Galler posted:

I checked one of our 690s and your bay counts are correct. However the 5.25 bays have metal dividers inside them so I don't believe you could put in a 3x4 converter without some demeling. I can post some pictures when I get home if you want.

What does demeling mean? Google turns up some place in Indonesia. I guess what you mean is some good old fashioned breakingmodifying poo poo? Like bending/cutting out the dividers?

Either way, I really appreciate you (and the other guy) looking into for a total random stranger on the internet.

I'm pretty sure I'm gonna land in possession of this computer/case. I will let you know how it goes!

deimos
Nov 30, 2006

Forget it man this bat is whack, it's got poobrain!

Gozinbulx posted:

What does demeling mean? Google turns up some place in Indonesia. I guess what you mean is some good old fashioned breakingmodifying poo poo? Like bending/cutting out the dividers?

Either way, I really appreciate you (and the other guy) looking into for a total random stranger on the internet.

I'm pretty sure I'm gonna land in possession of this computer/case. I will let you know how it goes!

pretty sure he missed an r in dremeling, the act of using a dremel to void warranties.

Traxxus
Jul 13, 2003

WWJD - What Would Jack Do?
My coolermaster had those, I just used a c-clamp and a screwdriver to bend them flat.

Galler
Jan 28, 2008


deimos posted:

pretty sure he missed an r in dremeling, the act of using a dremel to void warranties.

Yup. I was fighting with the iPad to not horribly mangle what I was typing and I lost.

McGlockenshire
Dec 16, 2005

GOLLOCKS!

Combat Pretzel posted:

Doesn't the Intel RST driver support hot swapping? I've enabled the specific port in the BIOS and slapped the Intel driver on the system. Yet I don't have the Safely Remove option. I've disabled the write cache on the port and can pull the drive at will, but I'd rather want to spin it down before pulling it.

It's a P67 based board, running Windows 8 and Intel RST 12.x.

--edit: Well, I've ditched the Intel driver and went back to Microsoft's driver. I get the remove option, but the drive doesn't spin down. I guess I'll keep the MS driver, because there seems evidence that it's slightly faster for SSD drives. I've also tried HotSwap!, the drive doesn't spin down. You can hear it start spinning down, just to go back to full spindle speed again.

I've never tried hot-swap under Windows, but my understanding is that it should work at least under Linux when RST is managed through mdraid.

Not all motherboards support hot-plugging as a native, published feature. I've only ever seen it in Q77 boards, and I haven't had the chance to test it there, and I don't know if you can actually get the drive to spin down.

Combat Pretzel
Jun 23, 2004

No, seriously... what kurds?!
It seems to work with Microsoft's driver. My main concern is that the drive's not being spun down when using safe remove. Based on the disks that I've hacked apart in the past, the heads should move to the safe zone or off the platter, depending on the model, when spinning down on power loss. But that's really just conjecture.

sharktamer
Oct 30, 2011

Shark tamer ridiculous
I'm looking to pick up an n40l, decided to go for it over the unraid solution to save some cash, especially if I can go for the rebate. It may struggle a bit with plex transcoding, especially with 1080p files, but it should work fine with simple serving.

I wanted to ask some questions about RAID. If I purchase the n40l and opt for just one 3tb drive now, will using RAID cause me issues if I plan on adding an extra drive in the future? It seems a bit silly to buy more storage than I need now, when I can buy extra when it's necessary and save in the long run. Is it a neccessarily evil to overshoot on storage though?

Galler
Jan 28, 2008


Gozinbulx posted:

Either way, I really appreciate you (and the other guy) looking into for a total random stranger on the internet.

I'm pretty sure I'm gonna land in possession of this computer/case. I will let you know how it goes!

Pics:




That inner rectangular bit at the bottom is a removable insert for the floppy drive. These cases are huge and if you were to swap out the motherboard and such you could fit drives all over the inside of it. If you keep the stock parts just be aware that it's pretty power hungry and not terribly quite.

poverty goat
Feb 15, 2004



Longinus00 posted:

It's not turn-key and it won't work with freenas but it's certainly possible.

http://multipath-tcp.org/pmwiki.php?n=Main.50Gbps

What about iSCSI MPIO? It sounds like it accomplishes the same thing and it's supported by freeNAS e: but I can't find a firm answer on whether it works for a single data stream or not

poverty goat fucked around with this message at 23:28 on Apr 13, 2013

Goon Matchmaker
Oct 23, 2003

I play too much EVE-Online
I use iSCSI MPIO on my VMWare environment at work. With two 1 gigabit links on the ESXi host and 2 gigabit links on the SAN we can get about 1.6Gbps from a single host. ESXi requries a bit of tweaking to get that working right. Mainly setting the pathing to round robin and adjusting the number of iops that go down each path before flipping from 1000 to 1.

Gozinbulx
Feb 19, 2004

Galler posted:

Pics:




That inner rectangular bit at the bottom is a removable insert for the floppy drive. These cases are huge and if you were to swap out the motherboard and such you could fit drives all over the inside of it. If you keep the stock parts just be aware that it's pretty power hungry and not terribly quite.

Thanks man, appreciate it!

Adbot
ADBOT LOVES YOU

fatman1683
Jan 8, 2004
.
How would one go about building an eSATA DAS? I'd like to have an external RAID box connected to my ESXi host, but the cheap off-the-shelf options don't have any kind of useful management interface, and there's a huge gap between the $200 models and the $2000+ models where there's basically nothing.

Presumably you'd need some kind of controller card in the storage box that can act as a disk and present the array in the box to the host's controller, but I have no idea if such a thing exists as a standalone product, and I haven't found anything that advertises that capability.

Anyone know if this is possible?

  • 1
  • 2
  • 3
  • 4
  • 5
  • Post
  • Reply