July 13, 201510 yr EDIT: Could someone move this to v6, sorry i was in the wrong board I was wondering what is the best place to start with what could be causing performance issues going from my cache to array. Normally, or prior to 6.0/6.0.1 final I would steadily (via Windows 8.1 file cut/paste) get about 50mb a sec avg. Now I get speeds all over the place and tons and tons of pausing to the point of copy failure. I know copying via my windows machine isn't ideal but it never had a problem before. My cache is a Samsung 850 SSD (256GB). My array is 15TB of random drives, about 11TB full. I am just kind of looking for things to look for in logs/
July 14, 201510 yr Disk 4 has a bad or loose SATA cable or a bad or dirty SATA port. Start in SAFE-mode and test performance.
July 14, 201510 yr Author Disk 4 has a bad or loose SATA cable or a bad or dirty SATA port. Start in SAFE-mode and test performance. I will give this a go tonight, just out of curiosity, what log pointed you to that?
July 15, 201510 yr Looking at your syslog, probably this part: Jul 13 18:54:44 Hades kernel: ata4.00: irq_stat 0x08000000, interface fatal error Jul 13 18:54:44 Hades kernel: ata4: SError: { UnrecovData HostInt 10B8B BadCRC } Jul 13 18:54:44 Hades kernel: ata4.00: failed command: READ DMA EXT Jul 13 18:54:44 Hades kernel: ata4.00: cmd 25/00:08:90:46:e4/00:00:75:00:00/e0 tag 8 dma 4096 in Jul 13 18:54:44 Hades kernel: res 50/00:00:2f:f4:9d/00:00:d2:00:00/e0 Emask 0x50 (ATA bus error) Jul 13 18:54:44 Hades kernel: ata4.00: status: { DRDY } Jul 13 18:54:44 Hades kernel: ata4: hard resetting link Jul 13 18:54:45 Hades kernel: ata4: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
August 18, 201510 yr Sorry to reply to an old topic, but was this resolved OP? I'm experiencing exactly the same issue, sadly I'm not seeing anything in my syslog... (I don't think). FYI: I'm passing through a Dell H310 (Reflashed as LSI). Logs attached. unraid-diagnostics-20150818-0959.zip
August 18, 201510 yr Community Expert Sorry to reply to an old topic, but was this resolved OP? I'm experiencing exactly the same issue, sadly I'm not seeing anything in my syslog... (I don't think). FYI: I'm passing through a Dell H310 (Reflashed as LSI). Logs attached. That diagnostic log seems to be from just after the array has been started. To help with diagnosing any problem you should include one which covers the period when the issue was being encountered.
August 18, 201510 yr Sorry to reply to an old topic, but was this resolved OP? I'm experiencing exactly the same issue, sadly I'm not seeing anything in my syslog... (I don't think). FYI: I'm passing through a Dell H310 (Reflashed as LSI). Logs attached. That diagnostic log seems to be from just after the array has been started. To help with diagnosing any problem you should include one which covers the period when the issue was being encountered. I thought it might have some insights, as looking through the log after it occurred this morning nothing at all was listed :-( I'll keep an eye out if the attached log doesn't contain any clues.
August 19, 201510 yr Update: The issue is fixed. Rolling the firmware on my LSI card from P20 to P19 resolved the issue!
December 21, 201510 yr I just want to chip in by saying that I have the same SSD disk as cache drive and experiencing 25-30MB/s when moving files to the array. Apart from the parity disk - attached directory to the mb controller, 6Gb - all of the disks in the array are attached to a SASLP-MV8 controller (.21 firmware version).
Archived
This topic is now archived and is closed to further replies.