Jump to content

Unraid crashes on Parity Check


tazire

Recommended Posts

Not much else to add other than the title. As stated a parity check starts then the server hard resets. I have to turn off the power to the PSU for a couple of seconds to reset the server and have it boot again. I checked the system logs and this is what appears.#

 

Sep  1 15:48:05 SERVER kernel: tun: Universal TUN/TAP device driver, 1.6
Sep  1 15:48:07 SERVER ntfs-3g[12935]: Incomplete multi-sector transfer: magic: 0x58444e49  size: 4096  usa_ofs: 40  usa_count: 3  data: 1637  usn: 1636: Input/output error
Sep  1 15:48:07 SERVER emhttpd: req (2): startState=STARTED&file=&csrf_token=****************&cmdNoCheck=Cancel
Sep  1 15:48:07 SERVER kernel: mdcmd (69): nocheck 
Sep  1 15:48:07 SERVER kernel: md: md_do_sync: got signal, exit...
Sep  1 15:48:07 SERVER kernel: md: recovery thread: completion status: -4
Sep  1 15:48:09 SERVER ntfs-3g[12935]: Incomplete multi-sector transfer: magic: 0x58444e49  size: 4096  usa_ofs: 40  usa_count: 3  data: 1637  usn: 1636: Input/output error
Sep  1 15:48:09 SERVER ntfs-3g[12935]: Incomplete multi-sector transfer: magic: 0x58444e49  size: 4096  usa_ofs: 40  usa_count: 3  data: 1637  usn: 1636: Input/output error
Sep  1 15:48:09 SERVER ntfs-3g[12935]: Incomplete multi-sector transfer: magic: 0x58444e49  size: 4096  usa_ofs: 40  usa_count: 3  data: 1637  usn: 1636: Input/output error
Sep  1 15:48:09 SERVER ntfs-3g[12935]: Incomplete multi-sector transfer: magic: 0x58444e49  size: 4096  usa_ofs: 40  usa_count: 3  data: 1637  usn: 1636: Input/output error
Sep  1 15:48:09 SERVER ntfs-3g[12935]: Corrupt index block signature: vcn 0 inode 437
Sep  1 15:48:09 SERVER ntfs-3g[12935]: Failed to find place for new entry: Input/output error
Sep  1 15:48:09 SERVER ntfs-3g[12935]: Failed to add entry to the index: Input/output error
Sep  1 15:48:31 SERVER emhttpd: req (3): csrf_token=****************&title=System+Log&cmd=%2FwebGui%2Fscripts%2Ftail_log&arg1=syslog
Sep  1 15:48:31 SERVER emhttpd: cmd: /usr/local/emhttp/plugins/dynamix/scripts/tail_log syslog
Sep  1 15:48:38 SERVER avahi-daemon[8936]: Leaving mDNS multicast group on interface eth0.IPv6 with address fe80::428d:5cff:fe76:fd66.
Sep  1 15:48:38 SERVER avahi-daemon[8936]: Joining mDNS multicast group on interface eth0.IPv6 with address fd82:33f6:bde0:1:428d:5cff:fe76:fd66.
Sep  1 15:48:38 SERVER avahi-daemon[8936]: Registering new address record for fd82:33f6:bde0:1:428d:5cff:fe76:fd66 on eth0.*.
Sep  1 15:48:38 SERVER avahi-daemon[8936]: Withdrawing address record for fe80::428d:5cff:fe76:fd66 on eth0.

 

I have my PSU connected to a UPS so im not concerned about a power surge having effected it. But is this possibly still a PSU issue? I only ask as i did some research and this solved the issue for a few people.

Link to comment

Thanks for the quick response. The only reason I wasnt sure if it was the PSU is that while im not parity checking I can transcode 4k media with no lock up what so ever.

 

nothing in that system log that would give it away? I wouldnt be too sure from looking at it myself?

 

Still there is never a bad time to put a better PSU in it I guess. Sure ill order one and change it out and fingers crossed that solves it. 

 

If anyone else has any possible solutions im all ears.

Link to comment
1 hour ago, tazire said:

The only reason I wasnt sure if it was the PSU is that while im not parity checking I can transcode 4k media with no lock up what so ever.

 

Running a parity check is in no way comparable to transcoding media. The former spins up all array disks and is heavy on disk I/O, while the latter is light on I/O and heavy on processing.

 

2 hours ago, tazire said:

nothing in that system log that would give it away?

 

You haven't posted a syslog. In the snippet above the errors are related to an NTFS formatted disk, which has no part in the on-going parity check, other than sharing a power supply and possibly a disk controller.

Link to comment
On 9/1/2018 at 6:19 PM, John_M said:

You haven't posted a syslog. In the snippet above the errors are related to an NTFS formatted disk, which has no part in the on-going parity check, other than sharing a power supply and possibly a disk controller.

 

I realised it wasnt a full entry. Its just those were the last entries before the system crashed and just shut  down. Either way I have the server running and just turned off parity check until i get the new PSU. Been running fine since without parity check. 

Link to comment

It was 100% the PSU. 500W corsair served its purpose on the lower drive number servers i had built. just pushed it a touch too far i guess. Went overkill on the new one with a 1000W platinum but not ever something I want to skimp on. particularly with the current system i have in place.

 

Anyway thanks for the responses. 

Link to comment

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...