64-bit jaunty has frequent filesystem corruption

Bug #372526 reported by quixote on 2009-05-06
This bug affects 1 person
Affects Status Importance Assigned to Milestone
Jaunty Jackalope Backports

Bug Description

I'm not enough of an expert to know why this problem is happening, whether it relates to backports or just plain jaunty, or which aspect of synaptic is involved. It seems serious enough to need some expert attention, so that's why the bug report. My system is a Core 2 Duo P8400 2.26GHz, 4GB RAM, Intel GMA X4500 graphics, running Jaunty and Intrepid in a dual boot setup.

I upgraded from 8.10 to 9.04, which went fine, and started installing other programs using Synaptic (Thunderbird, vlc, etc.). (At that point, it was all Jaunty.) The install failed, and when I rebooted I was told to run fsck. There were a few hundred bad inodes and god-knows-what-all. The system came back up. Things seemed okay, and I continued customizing. Same problem, only this time gdm blew up and I wound up having to do a clean install, which was and is a major pain.

I figured I had a failing hard drive, and replaced it. I started over, installed a few things with apt-get, No problem. Then I used Synaptic to install Firefox 3.5. Error messages about a lack of xulrunner-1.9 in /usr/bin. The program itself was there, but the link was missing. Also, was again told to run fsck on startup and had a few dozen bad inodes and whatnot. Once the system was back up, I put the xulrunner link in by hand. No more error messages. Then I tried Thunderbird again, and Synaptic again didn't put a link in /usr/bin, desktop integration was bad, and the program wouldn't start.

I used the command line dpkg to purge it and reinstalled with apt-get. No problems.

That's when I started to think Synaptic had been the source of all my problems. I've since installed using apt-get in terminal: vlc, gthumb, ufraw, which all have lists of dependencies as long as your arm, and they all installed without a hitch. No problems on bootup either.

Assuming this issue isn't somehow unique to me, I hope this helps track down what it is.

Update: May 8th. Update-manager wouldn't load. It flashed up briefly and vanished. Under Hardy, I seem to remember having a problem with gksudo. Could something related have resurfaced?

Also, I used Firefox 3.5b4. It froze up the whole system on more than one occasion, requiring a hard restart. And on those occasions I had the same bad inode mess that required running fsck before the partition would boot. I'm pretty sure both synaptic and update-manager had problems after a reboot without FF running, but I'm also pretty sure that sometimes it was running.

I've also installed Jaunty on a Dell Latitude D630 (horrible machine, I don't recommend it!) Intel Core Duo T9300, 2GB ram, and (I think!) ATI graphics. Jaunty is running without issues on that setup.

description: updated
summary: - synaptic results in bad installs
+ memory leaks? synaptic? results in bad installs

The same behavior is being reported on a couple of ubuntuforum threads:

 (64bit Jaunty) Frequent Filesystem Corruption

random lockups with disk corruption

summary: - memory leaks? synaptic? results in bad installs
+ 64-bit jaunty has frequent filesystem corruption
description: updated

I can confirm this issue on my hardware, and I'd be happy to post any logs that would be relevant to fixing this problem at the request of any developers. I'm not at all sure what is causing the issue (probably not synaptic, but who knows...). I usually notice it first with firefox (the google search bar often locks up, and when it does, it always means that I need yet another fsck). If it will help, here are my specs:

Operating System: 64-bit Jaunty Jackalope (9.04)
Kernel: 2.6.28-11-generic
Filesystem: Ext 3
Processor: Intel(R) Core(TM)2 Duo CPU P8400 @ 2.26GHz
Graphics chip: ATI Technologies Inc RV710 [Radeon HD 4350]

This is probably a duplicate of

I've had a similar kind of behaviour (ext3 getting corrupted over time, eventually package management getting messed up etc) on a Dell Latitude D830. According to observations in the above bug report, the issue appears to be related to ICH8/ICH9 chipset families. In my case, lspci gives
00:1f.1 IDE interface: Intel Corporation 82801HBM/HEM (ICH8M/ICH8M-E) IDE Controller (rev 02)
00:1f.2 IDE interface: Intel Corporation 82801HBM/HEM (ICH8M/ICH8M-E) SATA IDE Controller (rev 02)

To post a comment you must log in.
This report contains Public information  Edit
Everyone can see this information.

Other bug subscribers