September 9, 200916 yr I am looking for some assistance to figuring out what is wrong with this system.... I included the current syslog. (in 3 parts) I have noticed a variety of problems with my Unraid server - -The most aggravating is the inability to copy via network of several gigs of data from one drive to another.... I am writing directly to disk/folder. Example1 from disk7 to disk3 without loss of network connectivity, with both my Vista PC and IMAC with leopard. Both machines are fully up to date on patches / service packs etc... Example2- Loss of connectivity when copying from the Vista PC directly to the unraid server. Thank you for the assistance. In a previous post we identified an issue with a failing Maxtor drive and it has since been removed from the array. SPECS: Unraid 4.6 Beta6 Hardware. MSI P43 Neo3 Celeron 2 gig Ram Using onboar Network 8 Seagate SATA + 2 Seagate IDE (Brand MSI Model P43 Neo3-F Supported CPU CPU Socket Type LGA 775 CPU Type Core 2 Quad / Core 2 Extreme / Core 2 Duo FSB 1600(O.C)/1333MHz Chipsets North Bridge Intel P43 South Bridge Intel ICH10 Memory Number of Memory Slots 4×240pin Memory Standard DDR2 1066(O.C) / 800 Maximum Memory Supported 16GB Channel Supported Dual Channel Expansion Slots PCI Express 2.0 x16 1 PCI Express x1 1 PCI Slots 4 Storage Devices PATA 1 x ATA100 2 Dev. Max SATA 3Gb/s 8 Onboard Video Onboard Video Chipset None Onboard Audio Audio Chipset Realtek ALC888 Audio Channels 8 Channels Onboard LAN LAN Chipset Realtek 8111C Max LAN Speed 10/100/1000Mbps Rear Panel Ports PS/2 2 COM 1 LPT 1 USB 4 x USB 2.0 Audio Ports 6 Ports Onboard USB Onboard USB 4 x USB 2.0 connectors Physical Spec Form Factor ATX Dimensions 12.0" x 9.6" Power Pin 24 Pin ) Disk status Model / Temperature Size Free Reads Writes Errors parity ata-ST31000528AS_9VP02N0B 31°C 976,762,552 - 5,523 2,378,917 0 disk1 ata-ST3750640AS 35°C 732,574,552 47,046,416 1,761,632 41 0 disk2 ata-ST3750640AS 36°C 732,574,552 44,376,720 1,754,356 40 0 disk3 ata-ST3750640AS 38°C 732,574,552 696,145,420 1,655,172 111,831 0 disk4 ata-ST3750640AS 37°C 732,574,552 499,327,512 1,758,342 46 0 disk5 ata-ST3500320AS 31°C 488,386,552 48,025,640 1,060,006 41 0 disk6 ata-ST3500320AS 29°C 488,386,552 488,338,792 1,062,775 41 0 disk7 ata-ST3500641A 33°C 488,386,552 394,538,856 1,188,337 83 0 disk8 ata-ST3500641A 33°C 488,386,552 488,091,032 1,018,793 46 0 cache ata-ST3300622AS 32°C 293,036,152 246,504,020 36 40 0 Command area Started Stop will take the array off-line. Parity is valid. Check will start a Parity-Check.
September 9, 200916 yr Author Running reiserfsck on drives. reiserfsck --check started at Tue Sep 8 22:23:31 2009 ########### Replaying journal.. Reiserfs journal '/dev/md1' in blocks [18..8211]: 0 transactions replayed Checking internal tree..finished Comparing bitmaps..finished Checking Semantic tree: finished No corruptions found There are on the filesystem: Leaves 179850 Internal nodes 1105 Directories 1409 Other files 165287 Data block pointers 171187279 (0 of them are zero) Safe links 0 ########### reiserfsck finished at Tue Sep 8 23:12:05 2009 ########### reiserfsck --check started at Tue Sep 8 23:02:07 2009 ########### Replaying journal.. Reiserfs journal '/dev/md2' in blocks [18..8211]: 0 transactions replayed Checking internal tree..finished Comparing bitmaps..finished Checking Semantic tree: finished No corruptions found There are on the filesystem: Leaves 180204 Internal nodes 1109 Directories 1023 Other files 141307 Data block pointers 171855826 (1481 of them are zero) Safe links 0 ########### reiserfsck finished at Tue Sep 8 23:47:52 2009 ########### reiserfsck --check started at Tue Sep 8 23:24:06 2009 ########### Replaying journal.. Reiserfs journal '/dev/md3' in blocks [18..8211]: 0 transactions replayed Checking internal tree..finished Comparing bitmaps..finished Checking Semantic tree: finished No corruptions found There are on the filesystem: Leaves 9142 Internal nodes 55 Directories 36 Other files 5289 Data block pointers 9084286 (0 of them are zero) Safe links 0 ########### reiserfsck --check started at Tue Sep 8 23:32:43 2009 ########### Replaying journal.. Reiserfs journal '/dev/md4' in blocks [18..8211]: 0 transactions replayed Checking internal tree..finished Comparing bitmaps..finished Checking Semantic tree: finished No corruptions found There are on the filesystem: Leaves 61776 Internal nodes 378 Directories 9344 Other files 50306 Data block pointers 58244714 (8908 of them are zero) Safe links 0 ########### reiserfsck finished at Tue Sep 8 23:51:13 2009 ###########
September 9, 200916 yr Author reiserfsck --check started at Wed Sep 9 00:01:32 2009 ########### Replaying journal.. Reiserfs journal '/dev/md5' in blocks [18..8211]: 0 transactions replayed Checking internal tree..finished Comparing bitmaps..finished Checking Semantic tree: finished No corruptions found There are on the filesystem: Leaves 111809 Internal nodes 688 Directories 744 Other files 91692 Data block pointers 109965794 (0 of them are zero) Safe links 0 ########### reiserfsck finished at Wed Sep 9 00:22:02 2009 reiserfsck --check started at Wed Sep 9 00:14:42 2009 ########### Replaying journal.. Reiserfs journal '/dev/md6' in blocks [18..8211]: 0 transactions replayed Checking internal tree..finished Comparing bitmaps..finished Checking Semantic tree: finished No corruptions found There are on the filesystem: Leaves 1 Internal nodes 0 Directories 2 Other files 1 Data block pointers 2 (0 of them are zero) Safe links 0 ########### reiserfsck finished at Wed Sep 9 00:15:24 2009 reiserfsck --check started at Tue Sep 8 23:59:06 2009 ########### Replaying journal.. Reiserfs journal '/dev/md7' in blocks [18..8211]: 0 transactions replayed Checking internal tree..finished Comparing bitmaps..finished Checking Semantic tree: finished No corruptions found There are on the filesystem: Leaves 24176 Internal nodes 152 Directories 90 Other files 13778 Data block pointers 23436146 (10487 of them are zero) Safe links 0 ########### reiserfsck finished at Wed Sep 9 00:04:41 2009 ########### reiserfsck --check started at Wed Sep 9 00:00:59 2009 ########### Replaying journal.. Reiserfs journal '/dev/md8' in blocks [18..8211]: 0 transactions replayed Checking internal tree..finished Comparing bitmaps..finished Checking Semantic tree: finished No corruptions found There are on the filesystem: Leaves 68 Internal nodes 1 Directories 9 Other files 76 Data block pointers 61874 (0 of them are zero) Safe links 0 ########### reiserfsck finished at Wed Sep 9 00:02:39 2009 ###########
September 9, 200916 yr Author After running the reiserfsck on drives, did a reboot. New Syslog is attached. thanks for the assistance.
September 9, 200916 yr Author Attached SYSLOG0.txt from locked up system. Followed instruction in another post. cd cp /var/log/syslog /boot/syslog.txt killall smbd nmbd sync for disk in /mnt/disk* /mnt/cache do umount $disk done mdcmd stop
September 9, 200916 yr Author Current syslog from system with the new Network card. Unraid is currently rebuilding parity.... Please let me know if this looks like a better configuration, and if the hardware appears to be running properly with the new network card. Thanks.
September 10, 200916 yr Wow, you gave that poor system a real workout, assigning and un-assigning drives and rebuilding parity over and over! Apart from a problem with the network chipset, I saw no issues at all, with the system or the drives. I imagine you saw how much longer it took to do a parity build with the 2 PATA drives included, roughly 57MB/s with SATA drives only, roughly 31MB/s with all drives. That is a huge drop, but thankfully is only related to the parity check and build speeds, not to normal drive access performance. By the way, it is easier on you and me, if you zip the syslogs and attach the zip, instead of cutting them up. You have the NETDEV WATCHDOG, r8169, "transmit timed out" issue that others have, and it would explain the problems you have been seeing. My usual advice would be to disable the onboard Realtek NIC and install a good network card, such as an Intel PRO/1000 PCIe card. But Joe and someone else have found good advice about this problem, a workaround. I'll need to find it and pass it along, probably create a FAQ entry for it. Looks like you may have already figured this out ... more in a bit ...
September 10, 200916 yr You have added what appears to be a PCI network card with a slightly different Realtek chipset, hopefully different enough not to have the issue you were having. I see no problems, but this is only a short test, and time will tell if that took care of it. Many other users are using both of these Realtek chipsets without issue. At the end of the last syslog is a sudden jump of 5 hours in the system clock. I assume this was a timezone correction of 5 hours? I usually see other syslog messages related to a timezone change, but did not see them here.
September 11, 200916 yr Author Thank you RobJ, for your review of my information and your suggestions. You are correct I did update the time zone, but it is changing the time on the pc to an incorrect time. so I probably need to put the time zone in the opposite of the current 5 hours... I am Central Time zone. I would prefer to use the onboard NIC- but if I can't I will get another intel card because of its chipset. The PATA drives do drop parity speeds considerably. I have them in place for now to transfer data to the SATA's. I worked out the system by adding drives one at a time to see when the parity speed dropped, that is when I remenbered the 2 PATA drives. ********** Someting new to report. During that point (adding and reconfiguring the array) I had a scarry moment when the web interface reported multiple drives as non formated, and offered to have them formated before starting the parity. I wonder how that could happen since all the drives are formated, and most have data. This is one of multiple times I have had this abberation of "unformated" showing up in the web interface. ********** New Questions: 1. What is a good PCI or PCI 1x SATA card for the Unraid. I read many suggestions, but I want 4 ports if possible. Would a 4 port SATA card slow the parity as much as the 2 PATA drives, or would it have no impact in your opinion? Your suggestion of card (PCI 4 Port at low price- or a PCI 1x 4 ports/2 port) is appreciated. 2. Is 57MB/sc for parity a good speed? Would a faster processor and 4 gig ram improve the performance of the Unraid Server? Thank you again for your assistance.
September 11, 200916 yr Someting new to report. During that point (adding and reconfiguring the array) I had a scarry moment when the web interface reported multiple drives as non formated, and offered to have them formated before starting the parity. I wonder how that could happen since all the drives are formated, and most have data. This is one of multiple times I have had this abberation of "unformated" showing up in the web interface. ********** Please see the FAQ entry, "Why is a disk showing as Unformatted?" New Questions: 1. What is a good PCI or PCI 1x SATA card for the Unraid. I read many suggestions, but I want 4 ports if possible. Would a 4 port SATA card slow the parity as much as the 2 PATA drives, or would it have no impact in your opinion? Your suggestion of card (PCI 4 Port at low price- or a PCI 1x 4 ports/2 port) is appreciated. Please see the Hardware Compatibility, PCI SATA Controllers section. The best performance at lowest price would be any of the SiI3132-based cards (PCIe X1, 2 SATA300 ports, fully support PMP for the future, as low as $20). The best 4 port card is the Adaptec 1430SA (PCIe X4, 4 SATA300 ports, $95). Parity check performance is seriously impacted by multiple drives on the PCI buss, but you don't run parity checks that often. So long as you make sure the parity drive is connected to the fastest bus, not the PCI bus, individual drive performance will be fine, no matter where they are connected. 2. Is 57MB/sc for parity a good speed? Would a faster processor and 4 gig ram improve the performance of the Unraid Server? That is a pretty good speed, see the User Benchmarks for examples. Neither a faster CPU or more RAM will help, since you are primarily I/O bound. Faster busses and faster drives are the most important factors.
September 11, 200916 yr The best 4 port card is the Adaptec 1430SA (PCIe X4, 4 SATA300 ports, $95). I think the Rosewill RC-218 might be a better deal. http://lime-technology.com/forum/index.php?topic=4139.0 http://lime-technology.com/forum/index.php?topic=2820.0 Jim White posted that he swapped out the 1430sa's for these. I happen to like the nice price and the fact that two ports can be redirected to external SATA units. I just bought one myself, so I'll post when it comes in.
September 11, 200916 yr And now you have heard better advice, from voices with more personal experience with these cards than I do!
September 12, 200916 yr Author Thank you for all the great info. ** I think the Rosewill RC-218 might be a better deal. http://lime-technology.com/forum/index.php?topic=4139.0 http://lime-technology.com/forum/index.php?topic=2820.0 ** I had strongly researched that information, but found that I only have one PCI 1x slot. Its not a 4x or 8x, so I don't think it would work in that slot. What I have found for PCI1x are only a 2 SATA ports. I have found several plain PCI cards with 4 SATA ports. Hence the parity speed questions... As RobJ said, parity is not calculated often, so not a major issue, but at 12 drives a 700 min parity check makes it less likely I will perform them often.... How about these: http://www.newegg.com/Product/Product.aspx?Item=N82E16815124027 http://www.newegg.com/Product/Product.aspx?Item=N82E16816124008 http://www.newegg.com/Product/Product.aspx?Item=N82E16816132003 Are the SYBA cards any good?
September 17, 200916 yr How about these: http://www.newegg.com/Product/Product.aspx?Item=N82E16815124027 http://www.newegg.com/Product/Product.aspx?Item=N82E16816124008 http://www.newegg.com/Product/Product.aspx?Item=N82E16816132003 Are the SYBA cards any good? I have the first, works fine, but it looks like Newegg has deactivated it, they still have the low profile version. The third looks like the Rosewill equivalent. Monoprice has a similar SiI3132-based card (all are 2 SATA II ports, PCI Express X1).
Archived
This topic is now archived and is closed to further replies.