knarf0007 Posted May 9, 2017 Share Posted May 9, 2017 Hi, in the last time I had several read errors on some drives. The drives are relatively new and I'm more or less sure that there is no real drive issue. The smart reports are good as well. Anyway, here is the log of the last failure: May 9 20:39:48 UnRaid-Server kernel: ata2.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x0May 9 20:39:48 UnRaid-Server kernel: ata2.00: irq_stat 0x40000001May 9 20:39:48 UnRaid-Server kernel: ata2.00: failed command: READ DMA EXTMay 9 20:39:48 UnRaid-Server kernel: ata2.00: cmd 25/00:08:08:b0:4f/00:00:7f:01:00/e0 tag 22 dma 4096 inMay 9 20:39:48 UnRaid-Server kernel: res 51/40:08:08:b0:4f/00:00:7f:01:00/e0 Emask 0x9 (media error)May 9 20:39:48 UnRaid-Server kernel: ata2.00: status: { DRDY ERR }May 9 20:39:48 UnRaid-Server kernel: ata2.00: error: { UNC }May 9 20:39:48 UnRaid-Server kernel: ata2.00: configured for UDMA/133May 9 20:39:48 UnRaid-Server kernel: sd 3:0:0:0: [sdc] tag#22 UNKNOWN(0x2003) Result: hostbyte=0x00 driverbyte=0x08May 9 20:39:48 UnRaid-Server kernel: sd 3:0:0:0: [sdc] tag#22 Sense Key : 0x3 [current] May 9 20:39:48 UnRaid-Server kernel: sd 3:0:0:0: [sdc] tag#22 ASC=0x11 ASCQ=0x4 May 9 20:39:48 UnRaid-Server kernel: sd 3:0:0:0: [sdc] tag#22 CDB: opcode=0x88 88 00 00 00 00 01 7f 4f b0 08 00 00 00 08 00 00May 9 20:39:48 UnRaid-Server kernel: blk_update_request: I/O error, dev sdc, sector 6430896136May 9 20:39:48 UnRaid-Server kernel: md: disk11 read error, sector=6430896072May 9 20:39:48 UnRaid-Server kernel: ata2: EH complete I checked all cable connections twice, with no effect. Is there any idea, what's the cause of all this. Thanks in advance. Greetings Frank Link to comment
knarf0007 Posted May 9, 2017 Author Share Posted May 9, 2017 o.k., sorry! Diagnostics file attached. unraid-server-diagnostics-20170509-2134.zip Link to comment
JorgeB Posted May 9, 2017 Share Posted May 9, 2017 Rather strange, read errors on three different disks, all look like disk errors but SMART looks fine for all disks. All 3 are on the onboard controller, you could try and swap with 3 disks on the SASLP and see if errors repeat, running an extended SMART test on those disks is also a good idea. Link to comment
knarf0007 Posted May 9, 2017 Author Share Posted May 9, 2017 O.k. thanks, I'll try this. Link to comment
knarf0007 Posted May 22, 2017 Author Share Posted May 22, 2017 Hi, after I tried several actions (change from internal to SASLP controller and changing a lot of sata cable) the final solution was rather simple. I have 14 drives (2xcache, 1xparity, 11xdata) and therefor used a lot of "cheap" Y-power cables to power up all the drives. After a phase of random read errors of different drives, I recently had a total failure of one of these drives (it was missing totaly). After this, I checked the complete power cabling and replaced all this cheap Y-power cables with better cables (original power cables suitable for my power supply). Now everthing works smooth and fine! So, if anyone had these "dammed" ;-) cheap Y-power cables in his setup, I can only strongly suggest to remove them totaly. These "things" seems to react extremely sensitive to vibrations (fans, etc.) because of there bad mechanical capabilities. Cost me a lot of nerves and it was may be naive of me to think, I can build up such a complex setup with cheap consumer parts. greetings Frank Link to comment
Zonediver Posted May 22, 2017 Share Posted May 22, 2017 4 hours ago, knarf0007 said: Hi, after I tried several actions (change from internal to SASLP controller and changing a lot of sata cable) the final solution was rather simple. I have 14 drives (2xcache, 1xparity, 11xdata) and therefor used a lot of "cheap" Y-power cables to power up all the drives. After a phase of random read errors of different drives, I recently had a total failure of one of these drives (it was missing totaly). After this, I checked the complete power cabling and replaced all this cheap Y-power cables with better cables (original power cables suitable for my power supply). Now everthing works smooth and fine! So, if anyone had these "dammed" ;-) cheap Y-power cables in his setup, I can only strongly suggest to remove them totaly. These "things" seems to react extremely sensitive to vibrations (fans, etc.) because of there bad mechanical capabilities. Cost me a lot of nerves and it was may be naive of me to think, I can build up such a complex setup with cheap consumer parts. greetings Frank Can you post a Picture of this Cables and/or the SATA-Connector on the Cable? I had a similar problem in the past and changed all my cables to this connector type and all is fine now. FYI: This connectors have blade-contacts Link to comment
Recommended Posts
Archived
This topic is now archived and is closed to further replies.