wrong parity error count

Some background information:

I am verifying the C2SEE MB I go recently with 4.5.6 free version unRAID.  I had a dirty shutdown and when system came back, parity check kick in and start making correction in parity, immediately 32 errors were fixed. Because I was testing and did not care about data protection  so I stop this parity check. After that I did many data copying in and out as well as system reboot to exercise this system and unRAID still remembered there were 32 parity error at last parity check, this was correct so far.

Then I decide to bring in protection by unassigned and re-assigned parity disk, parity re-calculation kick in accordingly however when it finished unRAID still report there were 32 errors even though there is nothing reported in syslog that unRAID had made any parity correction this time. Apparently, there is a bug in unRAID that during un-assign/re-assign parity disk, unRAID did not cleanup historical data in parity error count. At step 3 in following syslog, i think unRAID should cleanup historical data associated with parity because parity disk is gone.

(Last checked on 11/4/2010 5:56:18 PM, finding 32 errors.)

I will let syslog speaks for itself.

Nov 4 11:27:42 TestTower kernel: mdcmd (1764): stop      <---------(1) stop unraid.

Nov 4 11:27:42 TestTower kernel: md1: stopping

Nov 4 11:27:42 TestTower kernel: md2: stopping

Nov 4 11:27:49 TestTower kernel: md: unRAID driver removed  <--------(2) restart unRAID

Nov 4 11:27:49 TestTower kernel: md: unRAID driver 0.95.4 installed

Nov 4 11:27:49 TestTower kernel: md: disk0 missing                  <-------- (3) System recognize parity disk is gone

Nov 4 11:27:49 TestTower kernel: md: import disk1: [8,32] (sdc) ST31500341AS 9VS150BR offset: 63 size: 1465138552

Nov 4 11:27:49 TestTower kernel: md: import disk2: [8,0] (sda) ST31000333AS 6TE0EPNV offset: 63 size: 976762552

Nov 4 11:27:53 TestTower kernel: mdcmd (10): start DISABLE_DISK

Nov 4 11:27:53 TestTower kernel: mdcmd (12): check

Nov 4 11:27:53 TestTower kernel: md: recovery thread woken up ...

Nov 4 11:27:53 TestTower kernel: md: recovery thread has nothing to resync        <--- (4) no parity disk, nothing need to be done

Nov 4 11:28:01 TestTower kernel: mdcmd (22): stop          <--- (5) stop unRAID again

Nov 4 11:28:07 TestTower kernel: md: unRAID driver removed  <--- (6) restart unRAID

Nov 4 11:28:07 TestTower kernel: md: unRAID driver 0.95.4 installed

Nov 4 11:28:07 TestTower kernel: md: import disk0: [8,16] (sdb) Hitachi HDS72202 JK1171B8GNNVJS offset: 63 size: 1953514552

Nov 4 11:28:07 TestTower kernel: md: disk0 replaced  <--- (7) system recognize new parity disk has been assigned

Nov 4 11:28:07 TestTower kernel: md: import disk1: [8,32] (sdc) ST31500341AS 9VS150BR offset: 63 size: 1465138552

Nov 4 11:28:07 TestTower kernel: md: import disk2: [8,0] (sda) ST31000333AS 6TE0EPNV offset: 63 size: 976762552

Nov 4 11:28:11 TestTower kernel: mdcmd (11): start RECON_DISK  <-- (8th) Kick off disk reconstruction, parity recalculation

Nov 4 11:28:12 TestTower kernel: mdcmd (13): check

Nov 4 11:28:12 TestTower kernel: md: recovery thread woken up ...

Nov 4 11:28:12 TestTower kernel: md: recovery thread syncing parity disk ...

Nov 4 11:28:12 TestTower kernel: md: using 2304k window, over a total of 1953514552 blocks

Nov 4 17:56:18 TestTower kernel: md: sync done. time=23285sec rate=83895K/sec

Nov 4 17:56:18 TestTower kernel: md: recovery thread sync completion status: 0  <--- (9) parity recalculation finished, no any correction was made in between