fitbrit

I'm going through this exact same situation right now with a few months old Hitachi 2TB disk12. I just precleared another disk since my disk3 is failing according to a SMART test. I'm now wondering whether I should use the precleared drive to back up the read-only disk12 before running reiserfsck in anything other than --check mode? --check (still running) says that The level of the node (0) is not correct, (1) expected the problem in the internal node occured (143196162), whole subtree is skipped. This is being repeated for several other (xxxxxxxxx) values.

I was about to move 620 GB of my cache drive. All was working well, then I looked at the Main page and the cache drive temp was reading 0 C, and I couldn't access it. I stopped and restarted the array, and the cache drive appeared unformatted. I rebooted the server and everything was okay again. I was able to write to the cache drive. So I initiated a manual move from Midnight Commander and it failed. Sure enough, the main page is showing a 0 C temp for the drive again. I'm going to reboot again and see what happens. Any suggestions other than, "Dude you should move stuff off the cache drive more frequently?"

I believe at least one of them is SATA I. They served me well with unRAID until I got the Supermicro 4xPCIe mini-SAS card. I'll part with these for $10 each, $20 shipped, or $15 for both, $30 shipped. Local pick-up in Montreal welcome. Prices Dropped!

I'm running 4.7 and I'm soooo tempted to try version 5. Patience... patience...

Got three as well. Assuming I get them, I'll replace two of my 530s with these. Then I'll sell the two 530s and third 540 and donate the money to some friends who're doing a 150+ mile bike ride for cancer research.

Thanks everyone. Everything is fine now. For some reason, the Mac was trying to connect to the server via AFP. Even more mysteriously, all was working well a few weeks ago, so I don't know what was different now; either AFP used to be supported in unraid, or the Mac changed its connection settings by itself.

Thanks for the tip... I did some reading about how I'd find out this info. It seems that the Mac may be trying to connect to the server using AFP. (The server IP number is prefaced with "AFP"). I'll try using an SMB prefix to see if that will work.

Definitely the Mac - the wife hasn't quite got into the habit of letting it sleep. Definitely the Server too, because I just changed some drives the other day.

I occasionally use my wife's Macbook to play video files for our toddler. The server used to show up just fine on the network on the MacBook, but no longer does. I recently upgraded to unraid 4.7, but have no way of knowing whether that was when the Macbook stopped seeing the server. The whole house is wired with Cat6 and all the PCs have no problems seeing the server at all. Any suggestions? Did 4.7 turn off some options by default that may affect how the server is seen by a networked Mac?

You can't run unraid at the same time as windows. I do however, use two 5 bay Sans Digital eSATA enclosures with my unraid server. They work just fine, and allow me to run 20 data drives, parity and a cache drive. In your case, you'd have to make a dedicated unraid server and connect the SD box to it via eSATA; the drives in the SD are seen as integral unraid disks. This assumes, of course, that you've cleared and formatted the drives via unraid ir add-on scripts.

I'm currently replacing my 1 TB drives with 2 TB ones. I have a bunch of 1.5 TB drives that will eventually yield to 3 TB disks. 2 TB -> 3TB just doesn't seem to be enough of a jump.

Great find!

I assume this will work with unRAID?

When this exact same symptom happened to me with this card, I found that I had the wrong cables too. They fit perfectly into the card and drives, but they were reverse breakouts. Fortunately NCIX.com replaced the cables free of charge because their original description of the reverse breakouts (wrong cable for this card) made them sound like they were really forward breakout cables. Once i had the right cables all was well. The clue was in the card's bios as already mentioned; the card clearly wasn't recognising any of the 6 drives attached to it.

Hi Jon Welcome to unRAID. I'm glad I and others on RFD could finally convince you to try it. Did you buy a licence? You'll need one to have three disks and a parity. Some other advice: Listen to whatever the experts like JoeL, BRiT, dgaschk among a few others (most are moderators) say. They've saved my ass and data a few times when I've been impatient!

Where in Canada are you?

Cool to know. I've been using alt+, which allows strings to be entered, with wild cards, so for example B* will select any file/folder beginning with 'B'.

I've had quite a few problems over the past little while with my server. Now, I think I can say they're solved for now. The problem boiled down to two simultaneous component failures in the long run, plus real drive failures occasionally. It was a dangerous mix, but I think I'll come out of it okay, just wiser. 1) The 4-pin molex to SATA power adaptor going to my parity drive. The problem was on the molex end of the adaptor. This was on the "dark side" of my Centurion 590 case, so it wasn't easily accessible, and indeed I'd forgotten that it was an adaptor at all. I guess it worked intermittently, but I've had way too many parity drives fail on me in recent months. I fixed the adaptor by pushing some of the pins in. Voila! A drive that was supposedly failed (and gave many errors upon writing to it) is suddenly deemed fine; data is rebuilding onto it. Unfortunately, I just spent $160 on a new PSU, but at least it's super efficient and has 12 SATA connectors so no more using adaptors! It'll arrive next week. 2) The flat eSATA cable on one of my Sans Digital 5-bay cases. I've posted about slow parity checks, rebuilds etc. a few times now. A couple of posts recommended that I check the cabling, but since this was an enclosure, the cabling to each drive wasn't accessible. The eSATA connections seemed secure enough. I bought a rounded cable and installed it. Voila! Data rebuilding is going much faster, and more importantly, consistently, without constant resetting of the SATA connections. My unMenu last 6 lines of syslog are finally showing nothing more malicious than a single duplicated thumbnails file. Before it was writing to the log all the time regarding problems with the ata7 controller (the Sans Digital enclosure). Thanks to JoeL and a multitude of others who've helped me; I'm especially thankful that unMENU exists! I'm sometimes rash and do things before asking about them. I even did a couple of initconfigs when I shouldn't have, but some judicious renaming of super.bak files (I think) got me back out of trouble. I always learn the consequences of what I'm doing, so that's a small mercy. And I've learned a few things about what to look for in syslogs. Anyway, the lesson here is that sometimes more than one thing may be wrong when things aren't right, and can compound diagnosis and fixing.

I ended up getting this "Gold", single-railed, fully moldular PSU: http://ncix.com/products/index.php?sku=53717&vpn=CMPSU-750AX%20A&manufacture=Corsair&promoid=1295 Completely over my budget, but I hope over the years, its efficiency will pay me back in the background.

Thanks, for the response - it was what I feared. Also, that's a great idea. One that I hadn't thought of because the 2 TB drive is what I used to copy the data off the two 1 TB drives in the first place! I'll try this. Cheers.

I had a missing parity disk. I bought a WD 2TB Black drive and put it into the parity slot. Nada. Nothing I did could get unRAID to even see this drive - it wasn't even seen in the BIOS/POST screen. I removed drive and made arrangements to RMA it. I replaced it with a new 2 TB Hitachi drive. On a whim I checked the WD Black in an eSATA dock on my Windows PC: Drive was recognised, formatted and written to just fine. I put the Black drive in my Sans Digital eSATA 5-bay enclosure, which is part of my unraid array: Drive was recognised. I removed the Hitachi drive from parity services and, keeping the WD drive where it was, made it my parity drive. No problems. I precleared the Hitachi drive and added it to the array - still in the tower where my parity drive usually is. I figured the whole WD Black drive not being recognised was a fluke, so I swapped the Hitachi and WD drives putting the drives where they ought to be. I didn't want the parity drive to be in an external box. Again, the WD Black drive wasn't recognised, no matter what I tried. I had to swap them back. Now my Hitachi drive (disk 13) is misbehaving - failing basically. The drive was running just fine and had a great SMART report. I'm beginning to suspect that the PSU doesn't have enough juice for 12 drives and a slew of fans (8 120mm fans, including the drive cages ones, plus the CPU fan). It's a three x 12V rail Antec 650W PSU. I've had trouble balancing the power loads before. Could it be that when I installed the higher power WD Black drive, it just didn't have the power to spin it up? The problems with the Hitachi drive may be as a result of some power connector changes I made, disrupting the balance on the 12V rails. I'll be buying a single 12V rail PSU I suppose, to eliminate PSU from the equation. Any recommendations for one that I might continue to use if I get a Norco 4220 case or 4224 case eventually?

Background I've had numerous problems over the past year or so, ever since i moved from 19 data drives to 20, in fact, or more specifically, going from 11 internal drives to 12 in my Centurion 590 case. Almost all my parity drives have failed and I at first blamed this on 2TB WD EADS drives being unreliable. However, upon testing these drives after removal 3 of 4 have been fine and have been working okay in my external enclosures as part of my array. I tried changing the parity drive's SATA cable and SATA slot. Now I think there may be a PSU problem, either with the drive's power connector or with the PSU itself not having enough juice for hte 12th drive in the case. Now, changing the PSU is tantamount to rebuilding the server, and I don't have time to do this for a little while. I also don't want to have 12 drives in my server case at the moment, but I have a failed drive that needs replacing. Cut to the chase I have a disabled 2TB drive in my tower and I need to replace it with another drive, but not in my tower - in one of my two Sans Digital 5-drive bays. These are also full of drives, but two 1TB drives have no data on them. The two 1 TB drives are empty of data, have the same amount of free space on them, and show up as being of exactly the same size in Unraid. I've tested media files from the missing disk at random and they play back fine, suggesting that my parity drive is able to rebuild the data on the fly just fine. No really, cut to the chase already! Question is: Can I zero out the two drives using this method while there's a missing disk. By my reckoning, doing so should leave the parity info intact. I could then remove these two drives and the missing disk? I'd replace one of the 1 TB drives with a new 2 TB in order to rebuild the data from the missing disk onto it. A problem I foresee is how to tell unRAID to trust everything in my new configuration and to start building data onto the new disk, when it may see three missing disks and only one replacement.

Hi Joe Yes I screwed up and accidentally typed cd /boot instead of cd /root the first time. However, when I fixed it (before your post - thanks again!), the above happened - see my edited post. So I've got everything blue-balled and haven't hit start yet.

Hi Joe Thanks for the response. Something weird happened. I did an initconfig. All disks are blue-balled. I tried mdcmd set invalidslot 99, but I don't get: cmdOper=set cmdResult=ok I just get the next command prompt.

I have a disk that showed up as not installed part way during a parity check, but I believe it was a power connector problem. I unassigned and reassigned it, and now it's blue. I believe parity info is fine, but have no problem redoing parity-sync afterwards. I'd just like to know how to proceed right now without having to rebuild the disk and get the whole array green again when I start it up. I believe all the data is there, I don't want to run an initconfig if that's the wrong thing to do. Help appreciated.