ChkDsk hangs

hopefully right thread ask this....

i have hardware raid 5 array made of 4 1tb seagate drives. total available space 2.7tb.

the operating system windows server 2008 standard edition , boot disk on drive. raid5 drive data.

recently happened , upon reboot of server system wanted run chkdsk against raid5 volume and it allowed continue chkdsk. got stage 2 of 3 and it output 2 lines of text follows:

correcting error in index $i30 file 286076.
correcting error in index $i30 file 286076.

and there stayed. machine allowed stay in state 3 days. out of desperation hard reset.

the server booted , asked chkdsk again not allowed. when os booted raid5 array checked consistency. took 4 days complete. quite long time server not doing @ other checking raid5 volume.

all data accessible. @ least best can determined without attempting copy files off raid5 volume.

chkdsk ran properties of volume once raid5 consistency check had finished. system scheduled chkdsk run next time system booted.

the short , long of chkdsk never finishes in reasonable amount of time. reasonable meaning days, not hours. granted volume large, there possibility chkdsk hangs?

to eliminate hardware issues, array moved set of hardware same os , same underlining driver architecture.

the raid 5 volume recognized without problem.

upon reboot, system wanted run chkdsk on volume. allowed , has been in state 2.5 days.

is there reasonable amount of time chkdsk should finish in? chances there index corruption chkdsk can't handle , hangs?

these systems running on intel e8400 dual core processor @ 3ghz 8gb of dram.

the raid5 volume full 175gb remaining free.

any guidance on appreciated.

thanks in advance.

---

ted s. antonakis

 

hello,

 

based on description that, after booting system, raid 5 volume checked consistency; , according "correcting error in index $130 file xxxx", seems related master file table corruption.

 

i firstly double check if can make sure 'physical disks' in raid working properly, hardware raid specific tools.

 

then, please try booting windows recovery environment , run chkdsk there see how goes? can boot setup media , click "repair computer" on install page, click "command prompt" on system recovery options page.

 

besides, "chkdsk runs every time reboot system" problem may occur if don't have enough free space on volume windows allocate space internal data structures. volume marked dirty each time restart system.

 

try following command see if volume marked dirty:

 

fsutil dirty query e:         (assuming checking e: drive and, if c: drive, should move virtual memory drive)

if volume dirty, try command "chkdsk e: /f /x".

 

additionally, sorry there not reasonable amount of time chkdsk should finish in. have seen took more 1 week finish. however, share general information on why stage 2 may take long time finish:

 

phase 2: checking indexes

================

 

during second pass, chkdsk displays message tells chkdsk verifying indexes , again displays percent completed, counting 0 100 percent. during phase, chkdsk examines each of indexes on volume.

 

indexes ntfs directories. "percent completed" chkdsk displays during phase percentage of total number of volume's directories have been checked. during pass, chkdsk examines each directory on volume, checking internal consistency , verifying every file , directory represented file record segment in mft referenced @ least 1 directory. chkdsk confirms every file or subdirectory referenced in directory exists valid file record segment in mft , checks circular directory references. finally, chkdsk confirms time stamps , file size information files up-to-date in directory listings files.

 

at end of phase, chkdsk has made sure there no "orphaned" files , directory listings legitimate files. orphaned file file there legitimate file record segment there no listing in directory. orphaned file can restored proper directory if directory still exists. if proper directory no longer exists, chkdsk creates directory in root directory , places file there. if chkdsk finds directory listings file record segments no longer in use, or file record segments in use not correspond file listed in directory, chkdsk removes directory entry file record segment. duration of stage 2 varies because amount of time required process directory closely tied number of files or subdirectories listed in directory. because of dependency, percent complete indicator might not advance smoothly during stage 2, though indicator continues advance large directories.

 

best regards,

chang yin

Windows Server  >  Windows Server General Forum