bwperez

Lastnight I upgraded from unraid 6.2.4 to 6.3.2. The upgrade went smoothly and everything appeared to work fine. I then proceeded to try pass through a new USB card and accidentally passed through the USB bus with the unraid usb. The system lost access to the USB drive and started throwing error messages. I rebooted to server and was able to re-enable the VM manager but I no longer had my cache drives displayed. The unraid server showed one of the two drives in my cache pool missing and the syslog made reference to sata failures. After replacing the SATA cables, power cables and moving the two cache drives to different SATA ports am finally able to get both drives to show up in the Unraid GUI. However, the drives go into readonly mode. Jun 19 00:05:47 BriansTower kernel: sas: sas_scsi_find_task: task 0xffff88203662d700 is done Jun 19 00:05:47 BriansTower kernel: sas: sas_eh_handle_sas_errors: task 0xffff88203662d700 is done Jun 19 00:05:47 BriansTower kernel: sas: ata8: end_device-4:1: cmd error handler Jun 19 00:05:47 BriansTower kernel: sas: ata7: end_device-4:0: dev error handler Jun 19 00:05:47 BriansTower kernel: sas: ata8: end_device-4:1: dev error handler Jun 19 00:05:47 BriansTower kernel: ata8.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x6 frozen Jun 19 00:05:47 BriansTower kernel: ata8.00: failed command: WRITE DMA EXT Jun 19 00:05:47 BriansTower kernel: ata8.00: cmd 35/00:20:e0:42:aa/00:00:31:00:00/e0 tag 5 dma 16384 out Jun 19 00:05:47 BriansTower kernel: res 40/00:01:00:00:00/00:00:00:00:00/00 Emask 0x4 (timeout) Jun 19 00:05:47 BriansTower kernel: ata8.00: status: { DRDY } Jun 19 00:05:47 BriansTower kernel: sas: ata9: end_device-4:2: dev error handler Jun 19 00:05:47 BriansTower kernel: ata8: hard resetting link Jun 19 00:05:48 BriansTower kernel: ata8.00: configured for UDMA/33 Jun 19 00:05:48 BriansTower kernel: sd 4:0:1:0: [sdi] tag#0 UNKNOWN(0x2003) Result: hostbyte=0x00 driverbyte=0x08 Jun 19 00:05:48 BriansTower kernel: sd 4:0:1:0: [sdi] tag#0 Sense Key : 0x5 [current] Jun 19 00:05:48 BriansTower kernel: sd 4:0:1:0: [sdi] tag#0 ASC=0x21 ASCQ=0x4 Jun 19 00:05:48 BriansTower kernel: sd 4:0:1:0: [sdi] tag#0 CDB: opcode=0x2a 2a 00 31 aa 42 e0 00 00 20 00 Jun 19 00:05:48 BriansTower kernel: blk_update_request: I/O error, dev sdi, sector 833241824 Jun 19 00:05:48 BriansTower kernel: BTRFS error (device sdi1): bdev /dev/sdi1 errs: wr 233, rd 0, flush 0, corrupt 0, gen 0 Jun 19 00:05:48 BriansTower kernel: ata8: EH complete Jun 19 00:05:48 BriansTower kernel: BTRFS: error (device sdi1) in btrfs_commit_transaction:2227: errno=-5 IO failure (Error while writing out transact ion) Jun 19 00:05:48 BriansTower kernel: sas: --- Exit sas_scsi_recover_host: busy: 0 failed: 1 tries: 1 Jun 19 00:05:48 BriansTower kernel: BTRFS info (device sdi1): forced readonly Jun 19 00:05:48 BriansTower kernel: BTRFS warning (device sdi1): Skipping commit of aborted transaction. Any ideas how to proceed? brianstower-diagnostics-20170619-0007.zip

That was it. I switched the sata cable and re-seated the power connectors. Balance completed and my VMs start without hanging up. Thank you!

I added a second cache drive to my array. The VMs froze and the system became unresponsive. I rebooted the server to restore access. The VMs are no longer accessible upon reboot. I tried removing the new cache drive but that resulted in the original drive showing unmountable. I reattached the new cache drive and rebooted and this is what shows. I seem to have lost all my VM images. btrfs filesystem show: Label: none uuid: bf7fd42c-678e-4e5f-9d70-ee7d17ec4738 Total devices 2 FS bytes used 243.77GiB devid 1 size 465.76GiB used 464.76GiB path /dev/sdk1 devid 2 size 894.25GiB used 246.74GiB path /dev/sde1 btrfs filesystem df: Data, RAID1: total=246.74GiB, used=85.80GiB Data, single: total=217.01GiB, used=157.74GiB System, single: total=4.00MiB, used=112.00KiB Metadata, single: total=1.01GiB, used=231.14MiB GlobalReserve, single: total=80.00MiB, used=0.00B edit - I removed the cache drives from the configuration and mounted the original 500G drive using unassigned drives plugin. I was able to copy off one vdisk image file but others are failing root@BriansTower:/mnt/disks/Samsung_SSD_850_EVO_500GB_S21HNXAG484067F/VMs/mc2# cp vdisk1.img /mnt/user/Storage/VMBackups1/mc2 cp: error reading 'vdisk1.img': Input/output error syslog has the following Jan 2 02:37:39 BriansTower kernel: scsi 5:0:0:0: Direct-Access ATA Samsung SSD 850 1B6Q PQ: 0 ANSI: 5 Jan 2 02:37:39 BriansTower kernel: sd 5:0:0:0: [sdk] 976773168 512-byte logical blocks: (500 GB/466 GiB) Jan 2 02:37:39 BriansTower kernel: sd 5:0:0:0: [sdk] Write Protect is off Jan 2 02:37:39 BriansTower kernel: sd 5:0:0:0: [sdk] Mode Sense: 00 3a 00 00 Jan 2 02:37:39 BriansTower kernel: sd 5:0:0:0: [sdk] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA Jan 2 02:37:39 BriansTower kernel: sd 5:0:0:0: Attached scsi generic sg4 type 0 Jan 2 02:37:39 BriansTower kernel: sdk: sdk1 Jan 2 02:37:39 BriansTower kernel: sd 5:0:0:0: [sdk] Attached SCSI disk Jan 2 02:42:08 BriansTower kernel: btrfs_dev_stat_print_on_error: 67 callbacks suppressed Jan 2 02:42:08 BriansTower kernel: BTRFS error (device sde1): bdev /dev/sdk1 errs: wr 53, rd 25, flush 0, corrupt 0, gen 0 I thought all that would happen when I added the disk to the cache pool is it would mirror the existing disk image. I did notice that my available storage on the original drive went up to 700G with the addition of the new 960g drive. The only image that I really would like to restore is the mc2 one above. Am I out of luck? -edit I rebooted the system and added the two cache disks back into the unraid config. This seems to have gotten past the unmounted state and kicks off the balance process. It gets about 30% in and btrfs balance status: No balance found on '/mnt/cache' I've attached the latest syslog brianstower-diagnostics-20170102-0043.zip brianstower-diagnostics-20170102-0313.zip

http://www.ebay.com/itm/181527808860?_trksid=p2060353.m1438.l2649&ssPageName=STRK%3AMEBIDX%3AIT I offered $4 and it was accepted

I see that others have the preclear plugin working with 6.2 beta but I'm having issues 6.2 beta19 edit: disregard. I see this is already reported

Running through the BIOS update (and the subsequent FRUSDRUPDATE) *will* reconfigure the BMC for your chassis configuration and allow the BMC to properly control the cooling domains and substantially slow the fans down. This is mandatory for proper operation. After the update fan noise is now very low when idling ill do some more testing next week to see how it does under load after I receive the stock heatsink that match the air duct. If things hold up these intel chasses with powersupply will prove to be a fantastic bargin. under $230 including shipping for case, redundant 750 ps, 8x3.5 hot swap drives, 2x heat sinks, air duct and 2 extra 8x2.5 hot swap bays. I don't see how you can beat it. I'm debating building a second one. Brian

The same seller also has the redundant 750 watt chassis with 8x3.5" HSBP (with and without redundant chassis fans). But yeah, I'm going to load this new box with lots of 2.5" drives (won't be using unRAID), this'll be a dedicated ESXi box. I purchased the same chassis and air duct. I also bought the 8x3.5 tray separately for $40 and swapped out the 2 2.5 drive trays. I agree with the hairdryer description of the noise. I've only booted into the bios so far but setting the acoustic levels hasn't improved things. My 2600cp is on an ancient bios so I'm hoping that will make a difference. If not I'll swap out fans and put a different powers supply in .

I've given up using AMD cards due to hdcp issues in Windows under virtualization. Opelec shouldn't have those issues but if you think you will create a windows VMware just stick with an Nvidia card. There may be newer or better cards but im using this msi 720gt with great success. http://www.newegg.com/Product/Product.aspx?Item=N82E16814127805 Single slot 2g and low power usage.

It never fails. Spend hours researching and looking for ways to do something and the simplest answer is provided in minutes. I tested in a spare system and found the drive was empty. Exactly what I needed to confirm. Thanks for the help!

I pulled an old 1TB drive from the "junk drawer" that shows it is a reiserfs filesystem. when I try to view the contents in Ubuntu after it is auto-mounted it throws a message "This location could not be displayed." You do ont have permissions necessary to view the contents..." I've confirmed it is a reiserfs drive: ubuntu@ubuntu:/dev$ sudo blkid /dev/sdc1 /dev/sdc1: UUID="92ce110f-0c4b-423d-9711-bd6bde92a431" TYPE="reiserfs" Is there any way I can confirm if it was a data drive or a parity drive? Brian

I wasn't blaming, just pointing out that a message that says the drive is mounted read only would be better than "share XXXXX is deleted" would help. Also, if ntfs isn't supported for the cache drive why did it allow it without notifying me. The drive was new to me and I wasn't even aware it was formatted ntfs. Raising clear errors make supporting a product easier for both the developer and user. Thanks again for the help. I wouldn't have seen the issue without it.

The drive was originally formatted ntfs which was the issue. I took it out and plugged it into my pc and confirmed I could write to it. I then destroyed the primary partition and added it back to the Unraid server. Now everything works. A more descriptive error or warning when adding an ntfs cache drive would be nice. Thanks for the help.

Here are the diagnostics edstower-diagnostics-20151205-0847.zip

When I attempt to create a cache only share unraid pops a message saying the share has been deleted. Creating a non-cache only drive works fine. Unraid 6.q.6 pro key 2 1TB drives + 256G SSD for cache. I've setup two other servers and I'm never had this type of issue. Any ideas?

IOMMU is supported and already working with Video and USB pass through on other VMs. The motherboard and CPU are GA-990FXA-UD7 / FX8350