December 31, 200916 yr Hi, I think I have some problems... My unraid server was working fine. Old config : 4 Gig ram, Gigabyte GA-MA780G-UD3H MB, 4 1T samsung drives, 4.5 final. I needed more space so I bought 2 2T Hitachi drives. I used the latest preclear script to test them in another old PC. I then installed them in the unraid server by unassigning the old parity drive, and placing a new 2T drive in the parity slot. I selected "yes, i'm sure" and started the parity sync. The 2 other drives (old parity and other 2T drive) are not yet assigned a slot (first need to do a parity sync). They are however fully connected (data and power cable). The parity sync went fine until about 60 to 70% (well past the 1T drives). Then I heard clicking sounds from the drives, and the website to configure unraid could not be reached anymore. After some time, I pulled the plug (the machine did not respond any more). After bootup I got the following syslog. Some things seem very strange to me, eg: Dec 31 14:44:35 Tower kernel: ata3: softreset failed (device not ready) Dec 31 14:44:35 Tower kernel: ata3: applying SB600 PMP SRST workaround and retrying * ? Dec 31 14:44:35 Tower kernel: ata1: SATA max UDMA/133 irq_stat 0x00400040, connection status changed irq 22 Dec 31 14:44:35 Tower kernel: ata2: SATA max UDMA/133 abar m1024@0xfe02f000 port 0xfe02f180 irq 22 Dec 31 14:44:35 Tower kernel: ata5.00: ATA-8: Hitachi HDS722020ALA330, JKAOA20N, max UDMA/133 Dec 31 14:44:35 Tower kernel: ata5.00: 3907029168 sectors, multi 0: LBA48 NCQ (depth 31/32) Dec 31 14:44:35 Tower kernel: ata5.00: configured for UDMA/133 * UDMA/133 for a SATA2 drive? Dec 31 14:44:35 Tower kernel: sd 6:0:0:0: [sdg] 8027789 512-byte logical blocks: (4.11 GB/3.82 GiB) Dec 31 14:44:35 Tower kernel: sd 6:0:0:0: [sdg] Write Protect is off * I only have 6 drives, I think this is the USB stick? Dec 31 14:45:20 Tower ntpd_initres[1389]: host name not found: pool.ntp.org Dec 31 14:45:20 Tower ntpd_initres[1389]: couldn't resolve `pool.ntp.org', giving up on it * I can ping pool.ntp.org from my windows box, will check dns... Is my MB toast? All 3 1T drives are green, and I can access them. All files seem to be ok. I'm now doing again a parity sync.
December 31, 200916 yr I can see that I also have this in my logfile.. Dec 31 14:44:35 Tower kernel: ata5.00: ATA-8: Hitachi HDS722020ALA330, JKAOA20N, max UDMA/133 Dec 31 14:44:35 Tower kernel: ata5.00: 3907029168 sectors, multi 0: LBA48 NCQ (depth 31/32) Dec 31 14:44:35 Tower kernel: ata5.00: configured for UDMA/133 * UDMA/133 for a SATA2 drive? I have posted my syslog, but so far no one have make any comment on my set-up... is in this post But, What I can see so have I not any issues http://lime-technology.com/forum/index.php?topic=4949.0
December 31, 200916 yr Author I can see that I also have this in my logfile.. Dec 31 14:44:35 Tower kernel: ata5.00: ATA-8: Hitachi HDS722020ALA330, JKAOA20N, max UDMA/133 Dec 31 14:44:35 Tower kernel: ata5.00: 3907029168 sectors, multi 0: LBA48 NCQ (depth 31/32) Dec 31 14:44:35 Tower kernel: ata5.00: configured for UDMA/133 * UDMA/133 for a SATA2 drive? I have posted my syslog, but so far no one have make any comment on my set-up... is in this post http://lime-technology.com/forum/index.php?topic=4949.0 Makes me a little bit more at ease with those syslog entry's. I don't think I got those with the 4.5 beta's, but I'm not sure. Just upgraded to 4.5 final and my server just worked, didn't check syslog then. Does anyone know if a parity sync is supposed to be slower than a parity check? I got 80 Mb/s with a check (4 1T drives), and now with the new parity drive I get about 58 Mb/s (goes into the 90 Mb/s when past the 1T).
December 31, 200916 yr Does anyone know if a parity sync is supposed to be slower than a parity check? I got 80 Mb/s with a check (4 1T drives), and now with the new parity drive I get about 58 Mb/s (goes into the 90 Mb/s when past the 1T). Writing to the drives is always slower than reading from them. Since you are writing the entire 2TB drive, it will be slower... It sounds like it is doing just fine. (Far better than the 12-13 MB/s I get on my older IDE based array) Joe L.
December 31, 200916 yr Author Does anyone know if a parity sync is supposed to be slower than a parity check? I got 80 Mb/s with a check (4 1T drives), and now with the new parity drive I get about 58 Mb/s (goes into the 90 Mb/s when past the 1T). Writing to the drives is always slower than reading from them. Since you are writing the entire 2TB drive, it will be slower... It sounds like it is doing just fine. (Far better than the 12-13 MB/s I get on my older IDE based array) Joe L. Thank you, seems logical. I just hope it completes the check without crashing again... No use for a fast machine if it crashes before getting at the finish .
December 31, 200916 yr Does anyone know if a parity sync is supposed to be slower than a parity check? I got 80 Mb/s with a check (4 1T drives), and now with the new parity drive I get about 58 Mb/s (goes into the 90 Mb/s when past the 1T). Writing to the drives is always slower than reading from them. Since you are writing the entire 2TB drive, it will be slower... It sounds like it is doing just fine. (Far better than the 12-13 MB/s I get on my older IDE based array) Joe L. Thank you, seems logical. I just hope it completes the check without crashing again... No use for a fast machine if it crashes before getting at the finish . You might want to open up a telnet window and do a tail -f /var/log/syslog command in it. At least you'll see the errors if it does crash. Joe L.
December 31, 200916 yr Author Does anyone know if a parity sync is supposed to be slower than a parity check? I got 80 Mb/s with a check (4 1T drives), and now with the new parity drive I get about 58 Mb/s (goes into the 90 Mb/s when past the 1T). Writing to the drives is always slower than reading from them. Since you are writing the entire 2TB drive, it will be slower... It sounds like it is doing just fine. (Far better than the 12-13 MB/s I get on my older IDE based array) Joe L. Thank you, seems logical. I just hope it completes the check without crashing again... No use for a fast machine if it crashes before getting at the finish . You might want to open up a telnet window and do a tail -f /var/log/syslog command in it. At least you'll see the errors if it does crash. Joe L. Will do, thanks for the tip!
January 1, 201016 yr Author An update. My unraid server now completed the parity sync, and I did also do a parity check after that. No errors were found. Maybe the first sync failed because cabling problems? I'm happy it works again....
January 12, 201016 yr Hi, I think I have some problems... My unraid server was working fine. Old config : 4 Gig ram, Gigabyte GA-MA780G-UD3H MB, 4 1T samsung drives, 4.5 final. I needed more space so I bought 2 2T Hitachi drives. I used the latest preclear script to test them in another old PC. I then installed them in the unraid server by unassigning the old parity drive, and placing a new 2T drive in the parity slot. I selected "yes, i'm sure" and started the parity sync. The 2 other drives (old parity and other 2T drive) are not yet assigned a slot (first need to do a parity sync). They are however fully connected (data and power cable). The parity sync went fine until about 60 to 70% (well past the 1T drives). Then I heard clicking sounds from the drives, and the website to configure unraid could not be reached anymore. After some time, I pulled the plug (the machine did not respond any more). After bootup I got the following syslog. Some things seem very strange to me, eg: .... Is my MB toast? All 3 1T drives are green, and I can access them. All files seem to be ok. I'm now doing again a parity sync. Your attached syslog looks fine, no obvious problems. Whatever happened would have been logged in the previous syslog. It is best to try and capture it immediately, at the physical console if necessary, so we can see what the system saw then, when the problem was occurring and before it is lost on reboot. Installing Powerdown will allow you to use the Ctrl-Alt-Del key sequence, and save the syslog before it powers down. Dec 31 14:44:35 Tower kernel: ata3: softreset failed (device not ready) Dec 31 14:44:35 Tower kernel: ata3: applying SB600 PMP SRST workaround and retrying * ? Completely normal for all users with motherboards based on the SB600 and SB700 chipsets. Dec 31 14:44:35 Tower kernel: ata1: SATA max UDMA/133 irq_stat 0x00400040, connection status changed irq 22 Dec 31 14:44:35 Tower kernel: ata2: SATA max UDMA/133 abar m1024@0xfe02f000 port 0xfe02f180 irq 22 Completely normal. Dec 31 14:44:35 Tower kernel: ata5.00: ATA-8: Hitachi HDS722020ALA330, JKAOA20N, max UDMA/133 Dec 31 14:44:35 Tower kernel: ata5.00: 3907029168 sectors, multi 0: LBA48 NCQ (depth 31/32) Dec 31 14:44:35 Tower kernel: ata5.00: configured for UDMA/133 * UDMA/133 for a SATA2 drive? Completely normal. Linux uses/emulates the ATA interfaces, even for SATA drives. The '133' corresponds to 133MB/s, which is faster than needed for all current SATA drives. Dec 31 14:44:35 Tower kernel: sd 6:0:0:0: [sdg] 8027789 512-byte logical blocks: (4.11 GB/3.82 GiB) Dec 31 14:44:35 Tower kernel: sd 6:0:0:0: [sdg] Write Protect is off * I only have 6 drives, I think this is the USB stick? Yes. Dec 31 14:45:20 Tower ntpd_initres[1389]: host name not found: pool.ntp.org Dec 31 14:45:20 Tower ntpd_initres[1389]: couldn't resolve `pool.ntp.org', giving up on it * I can ping pool.ntp.org from my windows box, will check dns... Your unRAID server also needs a DNS server, so make sure it is set on the Settings tab of your unRAID Web Management pages. Usually, just set it to a copy of whatever you have in the Gateway setting. I believe that unRAID itself sets it after a reboot or two.
Archived
This topic is now archived and is closed to further replies.