In reply to:

---

Tried the 2.2.17 kernel linked above, and it worked fine. Flashed the ext 3 kernel again, and the same boot log as above.

---

Huh. I don't know what to tell you. It would be nice if we had the linux IDE guru around to maybe help translate that. Mark, are you still here? It doesn't surprise me that there are bugs in that ext3 version, but that one shows itself in a weird way. I did a quick google search for some of that error and most of the results I found suggest dieing hardware.

I had planned to try to get ext3 a little more current when I have some time (maybe this weekend). Hopefully that will help.

-Mike

One possibility is a bad sector where the journal file is - ext2 wouldn't have a problem, because it wouldn't be touching that sector, but ext3 would.

In reply to:

---

One possibility is a bad sector where the journal file is - ext2 wouldn't have a problem, because it wouldn't be touching that sector, but ext3 would.

---

Sounds good in theory, but he doesn't even have a journal file set up at the moment if I understand correctly. I suppose it could still be looking for one when it tries to do the ext3 mount before falling back to ext2...

Well I've been up for a little over two weeks no with no issues. Today, however, while doing a moderately large sync emplode has decided to do a "Checking Media -- Checking disk integrity" ... for some reason this is taking a long time. It's a 10+30 gig configuration. I don't remember it ever taking more than 10 minutes to do this check previously. Currently I'm at the 15 minute mark. The weird thing is that the previous sync was normal, and I haven't crashed the player in a long time.

Do you think this has anything to do with ext3? And shouldn't it go super fast with the new ext3 savvy fsck?

Greg

In reply to:

---

Do you think this has anything to do with ext3? And shouldn't it go super fast with the new ext3 savvy fsck?

---

Actually an ext3 fsck won't be any faster. The thing is it should never have to do an fsck in the first place as it fixes any file system errors when the volume is first mounted. The only reason you are even getting an fsck is because the player software explicitly calls it. IIRC hijack has an option to disable the periodic fscks that you are seeing. I haven't actually tried it myself though.

FWIW, on my 40 Gig empeg fscks always seem to take at least a half hour. Did it eventually finish OK?

BTW, if anybody is wondering why I haven't touched this in a while it is mainly because I am in the middle of moving so haven't had much free time to work on it. I am still hoping to get it patched up a bit farther to (hopefully) fix Drakino's problem and the remount problem I have. I probably won't be able to get to it until next month though as I won't have internet access (sacrilegious!) for a bit after finishing the move into the new place.

-Mike

The reason for the sudden filesystem check is that you have not yet done this:


tune2fs -c0 -i0 /dev/hdXX (replace XX with appropriate device suffix for the filesystem in question, eg. /dev/hda5 ).

Ahh .. you're right, I haven't done that. I will do it now though, thanks!

Greg

Mike (Comb) is this already in your doc for setup?

tune2fs -c0 -i0 /dev/hdXX

Mike (Comb) is this already in your doc for setup?

tune2fs -c0 -i0 /dev/hdXX

It is mentioned briefly in the last paragraph of the website but I don't give details. I haven't checked if the version of tune2fs on the empeg is new enough to set that flag. Did it work OK for you? If so I will add it to the doc.

-Mike

The reason for the sudden filesystem check is that you have not yet done this:


tune2fs -c0 -i0 /dev/hdXX (replace XX with appropriate device suffix for the filesystem in question, eg. /dev/hda5 ).

It appears that there is an option in hijack to bypass that check as well, but I tried it and it didn't work for me (the drives are still checked after 20 or so mounts). Could it be broken or does it not like ext3?

Oh, and welcome back, how was the climbing?

-Mike

I unmounted the drives, ran that command and it said something about ext3 being updated, so I assume it worked like it should. At the very least it didn't die or complain.

greg

I unmounted the drives, ran that command and it said something about ext3 being updated, so I assume it worked like it should. At the very least it didn't die or complain.

Sounds good. I'll update my page and try it myself once I get my network put back together in my new apartment. Oh, by the way, that ext3 page will be down for at least a week starting sometime saturday afternoon while I wait to get internet access at the new place, so if anybody wants any files grab them now.

-Mike

I am still hoping to get it patched up a bit farther to (hopefully) fix Drakino's problem

I finally began seeing other signs of problems with my drive 1, so I have contacted support about it. It seems ext3 just had a way of seeing the future problem :-)

It's to the point now where I get garbled music playback at times from data being read off the drive wrong. I haven't run ext3 since I last posted. so thats not causing the problem as far as I can tell.