Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Topics - SloPoke

Pages: [1]
1
RAID-F Bug Reports / I regret ever having upgraded from RAID-F 2.0 to 2.1
« on: September 20, 2015, 07:35:22 pm »
RAID-F 2.0 was working great.  All scheduled jobs ran, I was able to restore deleted files, update, validate and quick-validate ran without problems.

Installed 2.1, had to recreate email and schedules and now email doesn't work and validate reports corrupted files, memory leaks, weirdness with file name displays, and my 9 hour weekly overnight validate locked up my system partway through.

So, I uninstalled 2.1 and put 2.0 back.  Now I can't get anything to run, it keeps reporting :

2015-09-20 17:15:51,977] INFO : Executing quick-validate...
[2015-09-20 17:15:51,977] INFO : [quick-validate] started at: Sun Sep 20 17:15:51 PDT 2015
[2015-09-20 17:15:52,026] ERROR: [quick-validate] error: NullPointerException[null]
java.lang.NullPointerException
   at com.tchegbe.lib.io.a.g.c.a(Unknown Source)
   at com.tchegbe.lib.io.a.g.c.<unknown>(Unknown Source)
   at com.tchegbe.lib.io.a.g.c.<unknown>(Unknown Source)
   at com.tchegbe.lib.io.a.e.x.a(Unknown Source)
   at com.tchegbe.lib.io.a.e.c.a.run(Unknown Source)
   at java.lang.Thread.run(Unknown Source)
[2015-09-20 17:15:52,026] INFO : [quick-validate] completed at: Sun Sep 20 17:15:52 PDT 2015
[2015-09-20 17:15:57,004] INFO : Sending email alert - message=[My SnapShot RAID][QUICK-VALIDATE] FAILURE
[2015-09-20 17:15:59,932] INFO : [My SnapShot RAID][QUICK-VALIDATE] FAILURE - emailed...

OK, let me do a bare metal restore of my boot drive from just before I upgraded to 2.1.  Should be exactly like I had never installed 2.1, but nope,  no luck, still getting java.lang.NullPointerException.

Looks like I'll have to recreate from scratch. This is an entire day wasted.

2
After upgrading to 2.1 from 2.0, re-entering my schedules and email settings, I find email notifications are no longer working. The Verify button on email settings works; I do get the test message, but FlexRAID no longer sends mail when a scheduled event completes.  No indication it tries to send email in the log file. Create Alerts On Success and On Failure are checked, but When Scheduled Only is not.

Also the log file has an odd device display for new files found and files that no longer exist:

During an Update:
[2015-09-18 08:05:36,515] WARN : D:0:0\<directory>\<file> no longer exist! Skipping...

During a quick validate:
[2015-09-18 08:00:04,559] WARN : Deleted: D:5:0\<directory>\<file>


3
Snapshot RAID / Some files always show as changed, even after update
« on: October 28, 2014, 12:42:21 am »
Running 2.0 final [2014-08-16] [Snapshot 1.4 stable / Storage Pool 1.0 stable / Real-Time 1.0 experimental] in Snapshot mode on Windows Server 2012 Essentials R2.

I download photos from my camera into folders in a DRU on the server.  I then run a FleXRAID SNAPSHOT update to update the parity disk.  A quick validate after shows everything to be healthy and up to date.

I then geotag my photos which of course modifies them.  The geotag program preserves the file date and time of the photos. I then run an update and FlexRaid reports however many files were geotagged as changed as expected.  I then run a quick validate and all of the geotagged photos STILL show up as changed. No matter how many times I run an update after geotagging the photos, they still show up as changed during a quick validate.

The only way to fix this is to rename all the folders the geotagged files are in, run an update, rename them back again, and then do another update.  Oddly enough the second update shows nothing has changed.

I presume this behavior comes from the geotagging program preserving the file dates and times of the photos, but can't figure out how come quick validate still shows them as changed even after an update.  I've run into this behavior with corrupted dates on files where the date is far in the future, but the dates all seem normal with the geotagged photos.

4
General Discussion / A FlexRAID success story!
« on: January 29, 2014, 12:16:43 pm »
 ;D I thought it might be a good idea to relate how, with the help of some forensic data recovery software and FlexRAID, I was able to restore 4 failed DRU's back to bit perfect condition.  Thank you Brahim!

I have a 9 DRU, 1 PPU snapshot FlexRAID setup on a Windows Server 2012 Essentials system  The DRU's are all basically just the ServerFolders directory tree on individual 3 TB drives, and contain all my videos, pictures, music, documents, plus downloaded TV shows and movies.  I also have a large external drive that I use for backup.  Since I can't backup that huge amount of data, only my videos, music, and pictures are backed up.  5 of the drives are formatted ReFS as well as the backup drive; the rest are NTFS. A RAID snapshot update is performed every morning just after a quick validate, and a full validate is done every Sunday.

Last weekend I wanted to start on moving to Windows Server 2012 Essentials R2, so I shut down my system, swapped the boot SSD drive for a spare one, and installed R2 on it.  At some point during my install my system hung so I hit the reset button.  But the install completed and everything appeared normal until I put the original SSD back and booted up WS2012E again.  To my horror the system came back up with the 4 of the 5 ReFS drives corrupted, including my backup drive. Trying to open any of the corrupted drives brought up the message "X:\ is not accessible. The volume repair was not successful." No CHKDSK for ReFS drives as they are supposed to be self-repairing and more reliable than NTFS formatted drives. Yeah, right. Great, I thought, the backup drive is gone and I have more DRU's gone than I have PPU's.  All my irreplaceable videos, pictures, and music lost.

However, I am happy to say that after over a week of recovery efforts, everything is back!

The first thing to do was to stop all scheduled FlexRAID jobs and try to salvage as much as I could off of the corrupted drives using partition recovery software.  There are only a few software products supporting ReFS: ReclaiMe, REMO Recover, and R-Tools Data Recovery are the ones I found.  I used R-Tools to scan and copy as much as I could from each of the corrupted drives to a spare 3TB drive, format the corrupted ReFS drive as NTFS, and copy the recovered files back.  This had to be done individually for each of the 5 drives and took days, as the drives were quite full and the two copy operations sometimes took 20 hours to complete. I even managed to recover most of the backup drive files and used that to restore missing files that could not be copied from the corrupt drives.

Next step was to run a full FlexRAID validate.  This pointed out changed files, corrupt files, deleted files, and new files on the reconstructed DRU's.  There were plenty, as not all of the salvaged files were intact and some files that had been deleted months ago were retrieved from the corrupted drives. I deleted all the new and corrupt files and, for each DRU, moved the changed files to another drive just in case they could not be recovered, and ran a FlexRAID recover on the DRU.  The FlexRAID log was filled with thousands of "performing checksum" messages on the existing files, I assume because the original file dates had not been recovered intact.  However, FlexRAID recovered all the missing files on each DRU even with missing files on the other DRU's!  It took about 8 or 9 hours for each DRU, but all the deleted files were brought back intact.

I ran a final validate: no changes or bitrot detected.  The last step was a full verify: every bit verified correctly.

I am very happy with this result. Money well spent on FlexRAID.

A word of warning though: do not trust your data to ReFS formatted drives.  I don't know what happened to cause the corruption, but I have never had that kind of catastrophic failure with NTFS formatted partitions.

5
RAID-F Bug Reports / Email notifications failing after 2.0u11a install
« on: April 03, 2013, 02:41:12 pm »
Ever since I installed the 11a release, email notifications have been failing. I have Java 7u17 32-bit installed on WS2012E.


[2013-04-02 15:33:14,965] ERROR: Sending the email to the following server failed : smtp.live.com:587
org.apache.commons.mail.EmailException: Sending the email to the following server failed : smtp.live.com:587
   at java.lang.Void.<unknown>(Unknown Source)
   at java.lang.Void.<unknown>(Unknown Source)
   at java.lang.Void.<unknown>(Unknown Source)
   at java.lang.Void.<unknown>(Unknown Source)
Caused by: java.lang.InternalError: JIT ERROR: (RT) couldn't generate OBJ directory name
   ... 4 more

6
I upgraded my WHS2011 system to WS2012E over the weekend and ran into a couple of problems with my FlexRAID pooled snapshot setup. While folders and files could at least be indexed on the pool volume in WHS2011, even though you'd have to rebuild the entire index from scratch whenever files were added or deleted otherwise they wouldn't be indexed, I find that in WS2012E the pool drive cannot be indexed at all.  I get this error in the event viewer:

Crawl could not be completed on content source <file:V:/>.
Context:  Application, SystemIndex Catalog
Details:
   Incorrect function.  (HRESULT : 0x80070001) (0x80070001)

And that's it.  You can't search for any files on the pool volume.

Another problem, possibly related. I've turned the media server on so that my media library would be accessible from the remote web access.  In WHS2011 this worked, in WS2012E, Pictures, Videos, and Music folders all show up but all have zero files if they're on the pool volume. You can drill down into the folders, but they all show as empty. Put them on a real hard drive and they will be visible.

7
RAID-F Bug Reports / Files added to pooled drive on Windows not indexed
« on: September 05, 2012, 12:10:35 am »
Files copied/deleted/moved on a pooled drive with Snapshot RAID on Windows Server aren't indexed automatically.  I have to log on to the server and force it to rebuild the entire index from scratch for the new/deleted/moved files to show up (or not show up) in searches.  Running most recent version of FlexRAID on WHS 2011.

8
Storage Pooling / Windows Search indexing not working on pooled drive
« on: July 01, 2012, 11:11:54 am »
For some reason Windows Search indexing is not noting changes to my pooled drive on WHS 2011.  I can delete and rebuild the index, but once it's built, changes to any files on the pooled drive do not trigger a re-index.  I can tell this is happening because when I'm looking at how many items are indexed in the indexing options, if I add a file to an indexed folder that is not on the pooled drive, the index item count increases by one.  If I add a file to a folder on the pooled drive, the index count doesn't change.  Plus searching for files that I've deleted long ago still shows them in the index.  The pooled drive is NTFS and not marked as removable.

9
Running Realtime RAID with pooling on a WHS 2011 box and the log has quite a few of these entries in them.

[2012-6-1 16:5:6.446924][4324]      [GetOperationResult] ViewOperationException: ErrorCode=2
[2012-6-1 16:5:6.446925][9636]      [GetOperationResult] ViewOperationException: ErrorCode=2
[2012-6-1 16:5:9.457947][1500]      [GetOperationResult] ViewOperationException: ErrorCode=2
[2012-6-1 16:5:9.457948][9228]      [GetOperationResult] ViewOperationException: ErrorCode=2
[2012-6-1 16:5:13.388794][2896]      [GetOperationResult] ViewOperationException: ErrorCode=2
[2012-6-1 16:5:13.388794][1500]      [GetOperationResult] ViewOperationException: ErrorCode=2
[2012-6-1 16:5:21.235397][9228]      [GetOperationResult] ViewOperationException: ErrorCode=2
[2012-6-1 16:5:21.235398][4324]      [GetOperationResult] ViewOperationException: ErrorCode=2

Should I be concerned about them?

10
Starting to work with realtime RAID cruise control on a WHS 2011 machine.  I created a pool of 3 drives: 2TB DRU1, 1TB DRU2, 2TB PPU. Initialized RAID and started Storage Pool.  So far, so good.  I then wanted to remove the 1TB DRU2 from the pool, so I went to Drive Manager, highlighted the drive and clicked Remove from Storage Pool.  The Storage Pool stopped and the task started with:

The [<[fast-contraction]>] task has successfully initiated... Process number: 3

and then the message

ERROR: Unable to delete trans cache! Size=0

After that I am unable to start the Storage Pool; I get the same 'ERROR: Unable to delete trans cache! Size=0' message. If I restart the FlexRAID service, I am able to start the Storage Pool but the drive I am trying to remove is still there and now I can't remove it at all because Remove button is disabled.


Pages: [1]