Chkdsk log formatter

[Neil Waldhauer]

I used the check disk log formatter, CHKLGJFS.EXE to format the log file for my JFS formatted disk. The version is 1.9.5 by D Azarewicz.

I don't know what it means, so I'm attaching it here. I only included the first few lines. OS/2 World does not permit the whole file to be uploaded. It's too big.

[Doug Bissett]

I have no idea what it means, but if you have a DFSEE subscription, I would ask Jan. DFSEE may give you more information, in a friendlier way too.

[Andi B.]

... error writing M to d:...

Sound very familiar to me. But can't remember now what it triggers. Maybe there's even a mantis ticket for that. Did you search for this? I know for sure we have discussed this during the last years but my memory is vanishing....

[Neil Waldhauer]

I notice that ArcaOS has the old IBM chklgjfs.exe. I wonder what is the difference?

[Lars]

The chkdsk log ("Unrecoverable error writing M to d:. CHKDSK CANNOT CONTINUE.") means that your JFS transaction log is broken which is really bad news because I think that is non-recoverable.

The "M" should really spell "Metadata" but this is some problem either with the tool or with the JFS transaction log.
There had been a bug report for this in either the AN or the eCS bugtracker.

[Lars]

Found the AN bug report:
https://mantis.arcanoae.com/view.php?id=168

[Neil Waldhauer]

Well, unlike the bug reporter above, the disk was still usable. But the system would restart unexpectedly, which was unacceptable.

I get from this that the disk was damaged, but OS/2 was able to ignore the error and continue. Perhaps this was the cause of the unexpected system restarts, but probably not.

[Lars]

It might be possible that it will still start. But you will no longer be able to run chkdsk to clear up any potential errors because chkdsk.exe will abort operation if it finds something wrong with the transaction log.
That's really unfortunate (I think it would be better if chkdsk completely wiped the transaction log if it gets corrupted).
But now you might have some lurking errors. I had the very same, including these unexpected system restart. I never realized there was something wrong with my JFS partition (because the transaction log was fine) until I finally ran chkdsk /F:3 from another partition.  A whole bunch of errors turned up and I had about 400 MB of crosslinked files (fortunately, that was basically a source file tree that I had cloned from a GITHUB repo which I could just reget).
But after that, the JFS partition was in an errorfree state.

The only thing I can suggest is that you save the files to another partition/media and reformat the JFS partition.