May 10, 201214 yr On my second server, I've been trying for a solid six weeks to increase its capacity using old disks I had lying around. The oldest having warranty expire in 2010, from purchase in 2007. I added a new Supermicro card, new cables (one had a bad connector and had to be replaced), a tight fit in this Azza 910 case for the two Supermicro 5 in 3 cages, etc. Luckily there's no data of consequence on there. You can see in the attachment I had a functioning array with eleven disks. I went to add a couple more 1tb FALS disks that were removed from my main machine. All hell broke loose. DISK_DSBL, red dot on parity, slow parity sync (parity was attached to the Supermicro card), etc. Just a cluster f*ck, as far as I'm concerned. After the incident today, I decided to start again (hello initconfig, I know you all too well) with only disks that are still in warranty and known to work. One weird thing, was after I canceled one of the slow (180KB/s) syncs, I ran hdparm -t and got this: root@Other:~# hdparm -t /dev/sdb /dev/sdb: Timing buffered disk reads: 398 MB in 3.01 seconds = 132.08 MB/sec root@Other:~# hdparm -t /dev/sdc /dev/sdc: Timing buffered disk reads: 190 MB in 3.03 seconds = 62.70 MB/sec root@Other:~# hdparm -t /dev/sdd /dev/sdd: Timing buffered disk reads: 330 MB in 3.01 seconds = 109.71 MB/sec root@Other:~# hdparm -t /dev/sde /dev/sde: Timing buffered disk reads: 316 MB in 3.01 seconds = 105.13 MB/sec root@Other:~# hdparm -t /dev/sdf /dev/sdf: Timing buffered disk reads: 212 MB in 3.01 seconds = 70.45 MB/sec root@Other:~# hdparm -t /dev/sdg /dev/sdg: Timing buffered disk reads: 238 MB in 3.02 seconds = 78.78 MB/sec root@Other:~# hdparm -t /dev/sdh /dev/sdh: Timing buffered disk reads: 216 MB in 3.01 seconds = 71.71 MB/sec root@Other:~# hdparm -t /dev/sdi /dev/sdi: Timing buffered disk reads: 346 MB in 3.01 seconds = 114.92 MB/sec root@Other:~# hdparm -t /dev/sdj /dev/sdj: Timing buffered disk reads: 336 MB in 3.02 seconds = 111.36 MB/sec root@Other:~# hdparm -t /dev/sdk /dev/sdk: Timing buffered disk reads: 352 MB in 3.01 seconds = 116.89 MB/sec root@Other:~# hdparm -t /dev/sdl /dev/sdl: Timing buffered disk reads: 350 MB in 3.00 seconds = 116.62 MB/sec root@Other:~# hdparm -t /dev/sdm /dev/sdm: Timing buffered disk reads: 240 MB in 3.01 seconds = 79.68 MB/sec The oldest disks read the slowest, and had no data on them. Just all messed up. So everything comes out of the case, I move a few disks from the Icy Dock carriers to Supermicro, and start fresh. I wanted to use the same disk for parity, but place it on a mobo slot. Yes, just because. I started with four data and one parity. Sync was cooking along at ~50,000 KB/s. What a relief. It completes, so I power down and add a disk. ONE disk. It's there with no errors. syslog looks fine. Format. Done. I wait ten minutes, repeat, and get the same result. I've got one more 1tb FALS to add, and am keeping another for a warm spare. With all of these 1tb disks, I ran ./preclear_disk.sh -A -c 2 /dev/sdX instead of the single run through like before. Seems fine to me: No SMART attributes are FAILING_NOW 0 sectors were pending re-allocation before the start of the preclear. 0 sectors were pending re-allocation after pre-read in cycle 1 of 2. 0 sectors were pending re-allocation after zero of disk in cycle 1 of 2. 0 sectors were pending re-allocation after post-read in cycle 1 of 2. 0 sectors were pending re-allocation after zero of disk in cycle 2 of 2. 0 sectors are pending re-allocation at the end of the preclear, the number of sectors pending re-allocation did not change. 0 sectors had been re-allocated before the start of the preclear. 0 sectors are re-allocated at the end of the preclear, the number of sectors re-allocated did not change. It's been a really aggravating and difficult time for me, but I think all the trouble is behind me now. Crossing my fingers, anyway. BIG thanks to lewcass for a modification of the My Slots Worksheet. He adjusted it for a 5 horizontal, 4 vertical, 5 horizontal orientation. Hopefully others can use it too. Thanks for reading!
May 16, 201214 yr Not to jump off topic but I've probably had over 30-40 computers in my time already, hundreds of hard drives and never once had a bad cable or connector. To tell you the truth I didn't think they existed until I keep reading about them on here.
May 16, 201214 yr All the problems I've had with drives being disabled have ended up as being PSU related. Even with a TX-850, I had problems because I was overloading either the power rail or the cables by hanging too many 7200rpm drives off one cable. I removed some drives from the cable and redistributed the power and haven't had any problems since. That was after checking connections and replacing every SATA cable. It was never a SATA data cable problem AFAICT.
May 16, 201214 yr Not to jump off topic but I've probably had over 30-40 computers in my time already, hundreds of hard drives and never once had a bad cable or connector. To tell you the truth I didn't think they existed until I keep reading about them on here. A desktop has only 1, 2 drives in most cases. You usually use the nice new cables that came with your drive or mobo and your PSU has free plugs on it. in these ghetto-rigged storage servers with 10-22 drives.. you tend to see lots of cheap "Y" power splitters. lots of sata cables and tons of connections. sometimes you might think a cable is fully seated and it is not, you might plug in a cable, then go to plug in something else and your fat wrist/hand bumps into it and dislodges it. (not calling any one physically fat BTW). there tends to be room for error. and questionable parts from questionable sources... Many times it is also PSU related... and then there are always those aging disks of questionable nature
May 16, 201214 yr Also, in desktop machines it would be very rare to access multiple disks at a time. Therefore cross-talk induced noise pickup from one un-shielded SATA cable to another is not likely.
May 16, 201214 yr Not to jump off topic but I've probably had over 30-40 computers in my time already, hundreds of hard drives and never once had a bad cable or connector. To tell you the truth I didn't think they existed until I keep reading about them on here. Sorry to pull more OT - I have a HDD that works exactly backwards as far as the activity light. It's in a carrier. Put another drive in the same carrier and it works fine. The reversed light HDD works perfectly fine in every other way. I've set up/used hundreds of HDDs over the years and had never seen that problem and didn't think it was possible.
May 16, 201214 yr Not to jump off topic but I've probably had over 30-40 computers in my time already, hundreds of hard drives and never once had a bad cable or connector. To tell you the truth I didn't think they existed until I keep reading about them on here. Sorry to pull more OT - I have a HDD that works exactly backwards as far as the activity light. It's in a carrier. Put another drive in the same carrier and it works fine. The reversed light HDD works perfectly fine in every other way. I've set up/used hundreds of HDDs over the years and had never seen that problem and didn't think it was possible. I'll bet one manufacturer patented the ability to show status by lighting a light when the disk was active. To avoid paying a royalty, they in turn patented a light that is on when idle.
Archived
This topic is now archived and is closed to further replies.