Separate names with a comma.
Discussion in 'SSDs & Data Storage' started by odditory, Jan 5, 2010.
Are you able to check the SMART data on that particular drive? Perhaps it is failing physically.
Glad you figured it out. Wouldn't hurt to do some more checks on the drives to make sure they're ok now. Of course by replacing the mobo and psu at once, we'll never know for certain what the culprit was.
That was my next question.
How do i run smart tests on the drives ??
Are you running Windows or Linux?
(it does show "something" when you select the drive in the Areca webgui but to me, that information always seemed "off"/weird)
I use Hard Disk Sentinel and it shows the smart data for the individual drives on the array.
I'm running windows 2016.
Is HD sentinel free?
They have a trial version which I believe is time limited, so should be adequate to get some smart data pulled. Don't remember if the tests are available during the trial period. https://www.hdsentinel.com/download.php
It looks like there's a GUI for smartmontools (smartctl) that works on Windows: https://gsmartcontrol.sourceforge.io/home/index.php/
Downside, you have to manually add each drive (it's simple but tedious)
The "device type" field is where you specify which "slot" it is (in screenshot, I added #6)
My CPU runs at 60c.
Is that a good thing?
I've never been able to get a solid answer for mine. Post your specific card, maybe someone can chime in.
60C is fine. Max temperature varies by card, but the lowest is 90C as far as I'm aware. 1882 series is good until 105C and 1883 can handle 115C.
1880ix for me.
Is that valid for every sensor though? My ARC-1883ix-16-2G has cpu temp, controller temp, and chip temp. The chip temp gets the hottest at 60c, but we haven't had summer yet either.
The Areca 188x series are designed to survive running in consumer/prosumer hardware which have just basic airflow. The same 1883 that runs at 135F in a desktop machine runs at 119F in a Prolliiant with its higher performance cooling.
So I'm going to be replacing my 12 - 3TB seagates to 8 - 8TB HGST Deskstar NAS Drives soon on my ARC 1882IX-16...
I'm still on V1.51 (2012-07-04) firmware. I tried upgrading to 1.54 but it didn't work. Do I have to go to 1.52, 1.53 first?
Are there any new Advanced Configurations setting for HGST drives with the new firmware?
I think i'm gonna upgrade the ram on the card, however i was unable to fine ddr2 compatibilty list.
Anyone happen to know which ddr2 dimms are supported?
I *think* it tops out at 4GB. I've actually got a dead 1880ix-24 with a 4GB DIMM in it. The card is dead, but I've got no reason to suspect there is anything wrong with the DIMM. It's an official Areca DIMM, so it should be guaranteed compatible.
For cards that take DDR2, they do in fact go to 4GB. Working 4GB DIMMs are pretty rare though (ECC unbuffered and ranks/layout matters). If you have that old of a card, you probably won't see much benefit anyway.
So which model/make should i look for?
Unfortunately I'm not sure anymore. You might find some info if you search this thread. The last cards that took DDR2 was the 1880 generation and with them being 7+ years old now, Areca has taken down the product pages for it. I honestly wouldn't worry about upgrading that much. You won't see much of a difference in performance.
Sorry, perhaps I was unclear. If you want a 4 GB stick of Areca-approved RAM, I've got one I'm not using. You pay shipping and it's yours.
Final got around to flashing the firmware on Monday. Found the read me from 1.52 that stated:
For Controllers with any firmware versions before 1.52,
(1) Update to firmware 1.52 and other files except 1.52 BIOS, and then reboot
(2) Update BIOS 1.52 ( that is because firmware prior to 1.52 won't recognize BIOS 1.52), then reboot.
Did that then updated to 1.54 after.
Trying to find more information SED Raid Sets before I swap to the new drives now. Nothing in the manuals...
I'm looking for a Rackmount case that would fit the Areca 1880ix-12 card I have,
Since the card is extremely long, (25cm), any suggestions?
Most Supermicro chassis should accommodate it (836 is probably what you want). I used to run one in a Norco 4220 as well.
Do you happen to know if Norco 4308 can accommodate the card?
I don't think it will. While there would be 2cm of clearance at the end for you to use right angle cables, I believe due to its height, it would hit the hot-swap bays. I would also be concerned about the lack of airflow for the PCIe slots (since it would act like a physical barrier due to its length).
Anyone with a 1280ML upgrade to 4GB of memory? I have 2GB now with a Kingston KVR533D2E4/2G stick. Found a 4GB stick that looked almost equivalent, Kingston KVR667D2E5K2/4G, but the 1280ML still reads 2GB of memory. Did I get the wrong stick?
[Edit] Crap, I think the /4G is referring to the kit of two sticks, not the single stick. Anyone have a model number of a 4GB stick that is compatible with the 1280ML?
Callout to ARC-1222 owners – I found an interesting bit of info saying that versions of the 1222 prior to 1.2B had a faulty power mosfet (see review from Brenton P. here), leading to early deaths.
I'm planning to buy a used card – does anybody know how the revision numbers work? I can see ARC1222-SAS VER:1.2 on the PCB but I'm unsure if the "B" would be printed there as well?
Anyone happen to know if the 1880ix is compatible with x8 raiser card?
Should have no issues with passive riser cards.
Sorry for the delay, was on vacation last week. I can't help you with a direct answer. I can tell you that my 1220 is version 1.2 from 07, and I was using it until last year when I upgraded to a larger card. It too says version 1.2 right on the board, and there wouldn't be room for a B. It looks like there isn't room for it to exist on the 1222 you posted either. All I can say really is my 1.2 lasted without issue all this time, good luck.
Thanks for checking. I inspected a few more photos and noticed that some 1.2 boards have an A sticker on the notch next to PCIe slot, the one I pulled the trigger on has a B. It's only a guess, but I figured I'd chance it. Maybe I'm right, maybe I'm wrong.
Dang I didn't consider a sticker. I have a C sticker on mine, between the ch3/4 and ch1/2 sata ports.
does anyone know why today, after half a year of unusual traffic, my raid 6 dropped a 10 TB disk?
I started writing on the disk and the disk was down.
2018-5-3 8:25:6 Controller#1(PCI) Enc#2 SLOT 08 Time Out Error
2018-5-3 8:25:11 Controller#1(PCI) SG-DATA-E Volume Degraded
2018-5-3 8:25:11 Controller#1(PCI) RAID6 RaidSet Degraded
2018-5-3 8:25:12 Controller#1(PCI) Enc#2 SLOT 08 Device Removed
Raid card : ARC-1882IX-24
Controller Name ARC-1882IX-24
Firmware Version V1.54 2017-08-11
BOOT ROM Version V1.54 2017-08-11
PL Firmware Version 220.127.116.11
Main Processor 800MHz PPC440 RevD1E
CPU ICache Size 32KBytes
CPU DCache Size 32KBytes/Write Back
CPU SCache Size 1024KBytes/Write Through
System Memory 8192MB/1200MHz/ECC
Thank you for help.
What's this unusual traffic you experienced for half a year? The disk timed out, so the raid did as it's supposed to and dumped the problem disk. Might want to do some smart tests on the disk, it could actually be failing.
I pulled the disk out of the server and tested it and everything was fine.
The disk array serves as a NAS storage
Can not the SAS / SATA cable be defective?
Trafic is aprox mean.. 30GB per day ?
Do you think I have the disk that the controller has evaluated as faulty to send a replacement for a replacement for a piece?
Sure the cable or connector could be defective, but unless you were bumping it around I doubt it. You could try popping it back in and add it back into the raid if you were willing, but disks shouldn't time out for no reason. Hopefully someone else can chime in regarding the timeout error, I've never actually had that happen to my arrays.
Not with the cables did nothing.
I returned the disk last night to a disk field and rebuild.
Now I runing test :
Volume State Checking
Volume Check Style Fix Parity, Scrub Bad Block
Errors Found 0
Volume Set Information
Volume Set Name SG-DATA-E
Volume Serial 42f5f50653536907
Raid Set Name RAID6
Volume Capacity 30000.0GB (0xDA4750000)
SCSI Ch/Id/Lun 0/0/1
Raid Level Raid 6
Stripe Size 128KBytes
Block Size 512Bytes
Member Disks 5
Cables (that are/have never been touched since installation and have operated properly for months/years) can and do go bad. I have had it happen with SATA straight and SAS fanout cables. Anything from microfractures due to vibration, oxidation and even the occasional poltergeist can contribute to the failure of cables and/or the connectors they contain. Mark the drive as spare and put it back in rotation if it passes SMART and maybe a few rounds of DBAN. See if it happens again.
Regarding Areca and TLER/ERC/CCTL .... well guys ... Since going with 12TB drives I just have endless problems - i.E. not working in the end.
I have multiple Arecas 1260, 1880 and 1882ix-24 (4 GB) .... The 1882 was even on a compatibility list from my HGST Ultrastar/Enterprise 12TBs .... but no TLER/ERC/TTL - Newest Firmware on Areca, Newest BIOS, Newest OS.
And on other drives (12TB Ironwolf (non Pro) ... TLER/ERC/CCTL works on 1260, but NOT1880 ... but 1882 works again ....
Areca support is somewhat unhelpful and just says basically.... well it recognizes the disk ....then it's fine ... I did explain the problem carefully. But Kevin doesn't seem to bother.
I did stress test them with HGST 12TB + Areca 1260 (with no TLER/ERC/CCTL) it did not take long and some HDs just dropped randomly out (no surprise).... The drives are brand new HGSTs and are okay..... I investes so much time in this BS and even the "both is on a compatibility list"helps nothing. Like proofed here. When you read the fine print they anyway don't do stress tests with most models - Then you also can drop the whole compatibility BS, since I never had a single SATA disk which was not recognized by any controller. And I have a few hunderd disks here.
Probably time to drop that Areca shit and move to ZFS.
I was very happy with Areca before with smaller/older disks, which I used heavily for years and did basically all you can do withit ... But it just doesn't work anymore.