Often repeated entry in iwtrace.log

What could this mean and why does it repeat roughly 60 times per SECOND (yes, 60 times a second, ~3600 times per minute):

Trying to delete non-existent vnode 0x.....

Dave

Current Environment(s):
(1) TS 6.5 on W2K3
(2) TS 6.1 SP1 on W2K3
By the way, I miss Unix terribly.

Find more posts tagged with

Comments

nipper

Don't see any errors like that in 6.5 with Sp1 on Unix.

You may want to iwfsck the backingstore. A vnode is a pointer that
the FS uses to map versions to Workareas, etc. Similar to an inode in Unix.

(I am certain my explanation is simplisitic & James or someone will correct/enhance).

BTW, typical caveats about iwfsck, just check, do not fix without talking to support first.

Good luck,

Andy

Migrateduser

Sorry, I forgot to mention the platform... and I typically ream others for doing the same thing. Yikes! This is on 6.5, but without SP1. I'll try iwfsck, though and will report any findings.

Dave

Current Environment(s):
(1) TS 6.5 on W2K3
(2) TS 6.1 SP1 on W2K3
By the way, I miss Unix terribly.

Migrateduser

No news is good news in this case, but the repeated message over 12 minutes time puzzles me. This occured from 9:05 - 9:17 this morning, which was before I got in. I also can't find anything in the other log files that point to any abnormalities at that time.


Command line:  c:\Interwoven\TeamSite\private\bin\iwfsck.exe
iwfsck: 6.5.0 Build 45607 Interwoven 20050208

***************************************
Walk branches and areas.
***************************************

[Mon May 02 12:37:49 2005]

***************************************
Walk directories, match with areas.
***************************************

[Mon May 02 12:37:49 2005]
[Mon May 02 12:37:49 2005]
Total points walked: 1578
Directories walked: 0
Files walked: 0
Symlinks walked: 0
TeamXML files walked: 0
Zombies walked: 0
Total points read from disk: 150
FINISHED 0 errors found

Dave

Current Environment(s):
(1) TS 6.5 on W2K3
(2) TS 6.1 SP1 on W2K3
By the way, I miss Unix terribly.

nipper

Well, since it is on Microsoft, you can reboot the server.

& hope it goes away.

Seriously you may want to log a support call, but with the iwfsck being OK
that is a big plus.

Andy

Migrateduser

I guess... I just implemented a log rotation system and it's a bit unnerving to have that particular log so large at this point in the day. But, it seems to be isolated, at least. Maybe if someone else reading this has seen this problem before, they can enlighten me? I haven't found anything on the support site or here on DevNet.

Dave

Current Environment(s):
(1) TS 6.5 on W2K3
(2) TS 6.1 SP1 on W2K3
By the way, I miss Unix terribly.

Cathrine

We are experiencing the same problems:

[Tue Apr 18 10:26:40 2006] TFSEnvironment: inode cache 100000
[Tue Apr 18 10:26:40 2006] Inode table was not shut down properly.
[Tue Apr 18 10:26:40 2006] Creating new inode tables.
[Tue Apr 18 10:26:42 2006]: TNetworkShares::Add Can't NetShareAdd, parameter error 0x8
[Tue Apr 18 10:26:42 2006]: TNetworkShares::Add Can't NetShareAdd, parameter error 0x8
[Tue Apr 18 10:26:42 2006] Batch Jobs Thread started
[Tue Apr 18 10:26:42 2006] c:\iw-home\etc\iw.cfg: rpc_threadcount=64
[Tue Apr 18 10:28:13 2006] Trying to delete non-existent vnode 0x835e6aa8

.... and then a whole series of "Trying to delete .."
I'm guessing it might have somehting to do with the "Inode table was not shut down properly", but I cant find anything about this in DevNet or support either. Anyone having similar problems?

chunk

We are. TeamSite lost its mind ... I mean iwmnt ... back on Apr 13 and again today. iwstat never returned a status nor an error. Had to kill -9 it to do anything. Our iw-store is on a SAN, but the connection was fine as we have another mount on the same SAN that worked fine.

iwtrace.log shows:
[Wed Apr 19 12:45:37 2006] Cache size: 400000
[Wed Apr 19 12:45:38 2006] TEnvironment: fd limit 65536 max 65536
[Wed Apr 19 12:45:38 2006] Initialized store 0x1 at "/opt/iw/iw-store/workflow" (version=300006, next id=0x1b8d00).
[Wed Apr 19 12:45:39 2006] TEnvironment: fd limit 65536 max 65536
[Wed Apr 19 12:45:39 2006] Initialized store 0x64 at "/opt/iw/iw-store/default" (version=300006, next id=0x302800).
[Wed Apr 19 12:45:39 2006] TFSEnvironment: inode cache 400000
[Wed Apr 19 12:45:39 2006] Inode table was not shut down properly.
[Wed Apr 19 12:45:39 2006] Creating new inode tables.
[Wed Apr 19 12:45:39 2006] Batch Jobs Thread started
[Wed Apr 19 12:45:39 2006] Started 8 FS threads
[Wed Apr 19 12:45:39 2006] Started Hopi manager on port 8898 with 256 threads
[Wed Apr 19 12:45:39 2006] /etc/iw.cfg: rpc_threadcount=64
[Wed Apr 19 12:45:39 2006] SCI Base Thread 285 started (max thread count 64)

Env:
TS 6.5 (no SP) on Sol 2.9

Box was running fine for roughly a eight months (reboots every 90 days). No OS patches in last 60 days.

Regards,

bill

rimcc

Hi,

We've been seeing this intermittently on our brand new Win2k3 EE, TS 6.5 SP2 box. It was somewhat annoying to have a beautifully clean setup, fresh installs etc. etc. and then have the logs fill with somewhat ominous looking messages.

We raised a support case, and the (summarised) response was:

[...]So far not able to determine the exact cause of this error in the iwtrace.log. 
[...]However, from the previous cases, it seems to be a harmless error
[...]suggested to be harmless and so ignored.

Apparently there is a bug filed for this (56991), although reading through the case again it seems this bug is more to do with making the message more descriptive, rather than fixing whatever is causing the messages themselves.

We raised our case back in Dec 2005 and haven't followed up on it for a while, so maybe there has been some progress on it since then.

Rich.

rimcc

Oh, and I meant to say: we've been living with the occasional burst of these messages for the past 4 months but, other than that weirdness, have experienced no instability. No disappearing Y: drive, no crashes, no hangs... and how much have I just jinxed myself?

Rich.

Migrateduser

We had a similar issue. The below description was the ultimate solution, although it was a pain to change so many configuration files.

"After fixing the store problem with you on the phone this morning I talked over what happened with the engineers. The problem started when you restored from backup into an existing store. This causes unrecoverable errors in the store because TeamSite can’t determine which on is valid. There is an old KB article on it (it’s for 4.5 on Solaris but the is issue/solution still apply) https://support.interwoven.com/kb/kb_show_article2.asp?ArticleID=2142

Stores are meant to be unique. If you wanted to recover one from backup, you would first need to delete the existing store by the same name. This doesn’t come up much since most users are restoring from backup because of a crash. In your case, you were just trying to overwrite the default store. In cases like yours, the best thing to do would have been to deactivate the store you want to over write. Then delete it and shut down all TeamSite services. Then restore from backup and reboot the server (if that is not possible then restart the TeamSite services manually). That should bring the new version of the store online without any problems."