TheR

Member
  • Content Count

    25
  • Joined

  • Last visited

Everything posted by TheR

  1. Another horror story? It looks I just don't have luck with WD Raptors, model WDC WD3000BLFS-0. I am running iSCSI storage server (Ubuntu 10.04) with 4 of them in Linux RAID10 on LSI on board controller and 4 of them in HW LSI raid controller. I had already upgraded firmware because of the problem with LSI raid controller described here Then there is this problem https://bugzilla.redhat.com/show_bug.cgi?id=512613 which is alive on 10.04 (althow I did't have the guts to try it for a year now). Now after more than 200 days uptime, last night software raid failed on me. <Log> Jun 6 20:24:54 ozssan kernel: [2279698.613730] mptbase: ioc0: LogInfo(0x31110b00): Originator={PL}, Code={Reset}, SubCode(0x0b00) Jun 6 20:24:54 ozssan kernel: [2279698.613918] mptbase: ioc0: LogInfo(0x31110b00): Originator={PL}, Code={Reset}, SubCode(0x0b00) Jun 6 20:24:54 ozssan kernel: [2279698.628061] mptbase: ioc0: LogInfo(0x30050000): Originator={IOP}, Code={Task Terminated}, SubCode(0x0000) Jun 6 20:24:54 ozssan kernel: [2279698.628069] sd 8:0:6:0: [sdd] Unhandled error code Jun 6 20:24:54 ozssan kernel: [2279698.628072] sd 8:0:6:0: [sdd] Result: hostbyte=DID_NO_CONNECT driverbyte=DRIVER_OK Jun 6 20:24:54 ozssan kernel: [2279698.628076] sd 8:0:6:0: [sdd] CDB: Read(10): 28 00 18 ad 55 bf 00 00 08 00 Jun 6 20:24:54 ozssan kernel: [2279698.628085] end_request: I/O error, dev sdd, sector 414012863 Jun 6 20:24:54 ozssan kernel: [2279698.636488] mptbase: ioc0: LogInfo(0x30050000): Originator={IOP}, Code={Task Terminated}, SubCode(0x0000) Jun 6 20:24:54 ozssan kernel: [2279698.636493] mptbase: ioc0: LogInfo(0x30050000): Originator={IOP}, Code={Task Terminated}, SubCode(0x0000) Jun 6 20:24:54 ozssan kernel: [2279698.636498] mptbase: ioc0: LogInfo(0x30050000): Originator={IOP}, Code={Task Terminated}, SubCode(0x0000) Jun 6 20:24:54 ozssan kernel: [2279698.636502] mptbase: ioc0: LogInfo(0x30050000): Originator={IOP}, Code={Task Terminated}, SubCode(0x0000) Jun 6 20:24:54 ozssan kernel: [2279698.636507] mptbase: ioc0: LogInfo(0x30050000): Originator={IOP}, Code={Task Terminated}, SubCode(0x0000) Jun 6 20:24:54 ozssan kernel: [2279698.636511] mptbase: ioc0: LogInfo(0x30050000): Originator={IOP}, Code={Task Terminated}, SubCode(0x0000) Jun 6 20:24:54 ozssan kernel: [2279698.636516] mptbase: ioc0: LogInfo(0x30050000): Originator={IOP}, Code={Task Terminated}, SubCode(0x0000) Jun 6 20:24:54 ozssan kernel: [2279698.636520] mptbase: ioc0: LogInfo(0x30050000): Originator={IOP}, Code={Task Terminated}, SubCode(0x0000) Jun 6 20:24:54 ozssan kernel: [2279698.636525] mptbase: ioc0: LogInfo(0x30050000): Originator={IOP}, Code={Task Terminated}, SubCode(0x0000) Jun 6 20:24:54 ozssan kernel: [2279698.636532] raid10: sdd1: rescheduling sector 828025600 Jun 6 20:24:54 ozssan kernel: [2279698.644896] sd 8:0:6:0: [sdd] Unhandled error code Jun 6 20:24:54 ozssan kernel: [2279698.644898] sd 8:0:6:0: [sdd] Result: hostbyte=DID_NO_CONNECT driverbyte=DRIVER_OK Jun 6 20:24:54 ozssan kernel: [2279698.644902] sd 8:0:6:0: [sdd] CDB: Read(10): 28 00 1c 9b 48 0f 00 00 30 00 Jun 6 20:24:54 ozssan kernel: [2279698.644910] end_request: I/O error, dev sdd, sector 479938575 Jun 6 20:24:54 ozssan kernel: [2279698.653200] raid10: sdd1: rescheduling sector 959876944 Jun 6 20:24:54 ozssan kernel: [2279698.661417] raid10: sdd1: rescheduling sector 959876952 Jun 6 20:24:54 ozssan kernel: [2279698.669461] raid10: sdd1: rescheduling sector 959876960 Jun 6 20:24:54 ozssan kernel: [2279698.677347] raid10: sdd1: rescheduling sector 959876968 Jun 6 20:24:54 ozssan kernel: [2279698.685027] raid10: sdd1: rescheduling sector 959876976 Jun 6 20:24:54 ozssan kernel: [2279698.692536] raid10: sdd1: rescheduling sector 959876984 Jun 6 20:24:54 ozssan kernel: [2279698.692540] sd 8:0:6:0: [sdd] Unhandled error code Jun 6 20:24:54 ozssan kernel: [2279698.692542] sd 8:0:6:0: [sdd] Result: hostbyte=DID_NO_CONNECT driverbyte=DRIVER_OK Jun 6 20:24:54 ozssan kernel: [2279698.692545] sd 8:0:6:0: [sdd] CDB: Write(10): 2a 00 00 29 7b 37 00 00 08 00 Jun 6 20:24:54 ozssan kernel: [2279698.692550] end_request: I/O error, dev sdd, sector 2718519 Jun 6 20:24:54 ozssan kernel: [2279698.692553] raid10: Disk failure on sdd1, disabling device. Jun 6 20:24:54 ozssan kernel: [2279698.692554] raid10: Operation continuing on 3 devices. </log> And after 8 seconds second disk failed and after another 20 seconds third disk failed too. Same log for all of drives. Pointless to say raid array was down. Interesting point is that 4'th drive didn't fail. But this drive has been replaced two months ago and has been powered up for about 60 days. Is there another ticking bomb in WD drives? by TheR
  2. TheR

    SSD Reliability

    This data kinda scares me. Aren't SSD's suppose to be more reliable than HDD's. And I am very surprised how much Sandforce disks die. Since Sandforce has some technology which should prevent loosing data if flash dies it looks like manufacturing problem or it is all a big bluff. by TheR
  3. It looks like Kingston is a way to go http://geizhals.at/eu/a415027.html by TheR
  4. Here in Europe you can get 4GB registered DIMM for less than 100€ but the same amount of unbuferred DIMM costs almost 400€. Is it "Only Macs are using it, so lets charge them more" or is production too smal or what. BOM should be smaller since you don't need extra buffer chip. On the other side 2GB UDIMMs are sold around 35€ while 4GB RDIMMs are around 45€. by TheR
  5. To answer to myself. It looks like there is very litle ddr3 demand for 4GB non-ECC modules too, since there are hard to find and also 3-4x more expensive than registrerred ones, which are probably on high demand for servers and therefor ''cheap''-er. I hate economy rules. We need 64bit only computers soon by TheR
  6. New Intel Corei7 procesors support 6 memory stick, which can be 4GB each for total of 24GB RAM. Which is quite a lot. And indeed some motherboard makers already have products that claim support for 24GB of RAM. Memory is DDR3 ECC, but not registered. I did some research and found that registered memory is mainly neeaded by servers with big amounts of memory, because memory controller would not be able to work with larger amounts of memory. My question is. Might Corei7 have some problems using this large amount of non buffred memory? by TheR
  7. This is a good information. It looks like will have to wait for (probably) more expensive Xeon version of CPU , because ECC should be essential for 24/7 machine. by TheR
  8. It is official. Promise's PCIe RAID controller. http://www.promise.com/product/product_det...&product_id=156 What do you think? Will it work on SLI board (2x PCIe x8) althow it is PCIe x4 card. I'am thinking AMD X2 4400+, decent SLI board (ASUS), 64bit Linux, 2TB of disk space. by TheR
  9. Years back I remember there was quite performance penalty when going from 2 DIMMS to 4 DIMMS RAM configuraction. I don't remember why was it so, but is this still a case with moderen chipsets and DDR2 memory. by TheR
  10. I just want to inform you that driver for EX8350 is included in Ubuntu 7.04. The card was recognized and installed automaticaly and I was able to access 2T partition instantly. It looks like driver is now part of Linux kernel. Festy is based on 2.6.20. by TheR
  11. TheR

    SSD's = Defrag is a thing of the past?

    He, he. SSD is fragmented by default , since it's logic tries to spread writes trough the available space because flash cell can be written only about million times before it dies. by TheR
  12. Can anybody confirm if Promise SuperTrak EX8350 driver is included in Linux Kernel 2.6.18 and above and therefore will be available in upcoming OpenSuse 10.2 Thank you TheR
  13. I am running EX8350 with more or less no problems on SuSE 10.0. More problems is of course when kernel gets updated . Speeds are around 50MB/s on 8 disk RAID 5 array. What I would like to know is which Linux distros do you run EX8350 on. Promise officialy supports only SuSE and RedHat with drivers so far and I am really interested changing to Ubuntu 606. by TheR
  14. SATA works too. Tyan Tomcat MB. I am stil trying to compile the drivers but no result at all. In a Linux kernel newsgroup http://marc.theaimsgroup.com/?l=linux-kern...28988110335&w=2 is a thread saying that drivers will be included in the kernel. But in 2.6.16.1 are still not there. How to compile driver for Suse Linux? by TheR
  15. <quote> #insmod shasta.ko Killed #lsmod Module Size Used by <the command hangs here> The output of dmesg is; CODE shasta: no version for "struct_module" found: kernel tainted. shasta: module not supported by Novell, setting U taint flag. Unable to handle kernel paging request at ffffffff8899bec0 RIP: <ffffffff80159593>{load_module+2851} PGD 103027 PUD 105027 PMD 7d61e067 PTE 0 Oops: 0002 [1] SMP CPU 1 etc....... </quote> Geronimo4711 were you able to resolve your problem and how? by TheR
  16. TheR

    Loading OS onto RAID 5 array?

    On server I would not prefer instaling OS on RAID5 drive, because if something fails and OS has to be reinstaled it can be quite messy to install RAID5 drivers if they are not already supported by OS. by TheR
  17. Can a LTO2 tape be used in a LTO1 drive? I am using Backup Exec 9.1 program for backup and the tape is recognized as ''Cleaning tape'' and can not be assigend to other backup sets. Only then rised the above question to me. Thanks for the answer TheR
  18. TheR

    What to choose?

    Looks like BIOS autodetection problem. Can you disable hard drive autodetection. by TheR
  19. Excuse me. Were you able to run card on TYAN board? You write that you will try, had some problems with drivers, but then nothing else. Could you tell more about your problems? by TheR
  20. If you ever need second GB ethernet on PCI it is a CPU hog. MSI Platinum series MB has onboard second ethernet chip which is connected through PCIe. I don't know is there any issue running PCIe 4x card in 2x slot as I don't know is there any probles running PCIe 4x card in PCIe 16x slot. 500MB/s (2x) of bandwith should be enough fro any disk operation. by TheR
  21. ASUS is no more an option, because its second network controller is connected via PCI. I've also read that althow some consumer motherboards have PCIe 4x slots mounted on motherboard, slot can only work in 2x mode. And I am not interested in graphics. I need a backup server which will also be used for other usefull functions. by TheR
  22. Can you recommend a good chassis for disk array. I am thinking of this http://www.supermicro.com/products/chassis...SC933T-R760.cfm althow it is not cheap. But if I count other components it will make about 20% of total price for 2TB array. Would there be any problem if I put ATX board into it? by TheR
  23. I just ordered one from mwave.com (backordered to 2005-08-26, though). I plan to run it on an NForce4 board (Tyan Tomcat K8E) under Linux 2.6 (Debian stable, amd64); we'll see how it goes.. at $380 it's half the price of the equivalent Areca, definitely worth a shot though I hate being the guinea pig. 213673[/snapback] I am thinking exactly the same configuration with Suse Linux. I have seen on Suse web site that the driver has got their Enterprise Server certificate. I am realy looking forward to your experience. by TheR
  24. I found it here. http://www.geizhals.at/eu/a159536.html Althow non of the sites has it on stock. by TheR
  25. TheR

    Installing SP2 results in:

    Checked that already. There's nothing listed in the Add/Remove applet 212549[/snapback] My problem turned to be ZoneAlarm firewall. I could post into safe mode, deinstall firewall and it was good. I would suggest to update all drivers prior to SP2 install. by TheR