For fsck sake
“Why is this remote restart taking so long?” I wondered, watching a successive stream of Request timed out messages being returned from PING.
One long walk to the server room later…
GAAAAH CTRL+C
Say what you want about Windows, but Windows Server doesn’t spontaneously decide to do a disk check during restart.
13 Responses to “For fsck sake”
Leave a Reply
Feed your Anger
AT on TechNet
- Helped identify a helpful topic WSUS - Declining Superseded Updates in the WSUS Forum. 2nd May, 2013
- Contributed a helpful post to the IE10 on Windows 7: Spellcheck is broken if AppData is redirected to a UNC path thread in the Internet Explorer 8, 9, 10 Forum. 1st March, 2013
- Asked the question IE10 on Windows 7: Spellcheck is broken if AppData is redirected to a UNC path in the Internet Explorer 8, 9, 10 Forum. 1st March, 2013
Copyright
"The Angry Technician" by angrytechnician.wordpress.com is licensed under a Creative Commons Attribution-Noncommercial-Share Alike 2.0 UK: England & Wales License.
Please read my submission policy.
True enough. We’ve actually disabled the periodic diskcheck in later installs. If you’d like to do this…
on the commandline…
mount
to show you mounted disks, look for /var/log and /, these should be /dev/sdaX or similar
then… for each one
tune2fs -c 0 /dev/sdaX
tune2fs -i 0 /dev/sdaX
and it will not check again unless it thinks something is wonky
No KVM over IP, or HP iLO? Lifesaver for us as our data center is in a different building than my office.
We’re a school with 1 server rack. Not really in our price range.
Could run it as a VM. Then you can view the console anywhere.
The server in question is actually an appliance; it has custom hardware that we have a support contract for. Smoothwall do offer the system as a VM now, but that’s not an option for us.
>> Windows Server doesn’t spontaneously decide to do a disk check during restart
Really? I don’t know what Windows Server you’ve been using, but mine likes to drop into that grey pre-boot bit and tell me I must press any key now to stop AUTOCHK whenever it bloody well feels like it.
There’s nothing wrong with a good fsck every 214 days. You should probably cron some read-only ones once a month, or so.
At least it didn’t hit an overflow bug in the kernel timer routines and panic after 208.5 days, like most of mine did last week…
Or, at 214 days, maybe it did…?
Yes , it only decides to spontaneously restart.
ZFS. That is all.
Mark I would love to agree however as Oracle support told me for the ZFS storage appliances we have that my poor performance was down to significant fragmentation, yes that is on my ZFS box so I was puzzled as I was told they don’t suffer etc…. yet these become completly unusable once badly fragmented to the point its wipe and rebuild from scratch to get them going :( Give me a windows box any day!
How long have you been wating for your UNIX system to have a disk check problem, so that you could use…
“What the FSCK” or similar.
Be honest!
Wallace
It was a while. We only have 2 *nix systems, and they usually only reboot in the small hours so I never notice if it takes ages.
For over a year, new installs use ext4 instead of ext3. There has been no in-place upgrade mechanism to do all partitions, but a recent update changes the OS partition to ext4. ext4 is about 20 times faster to fsck so the problem becomes moot. It is possible to in-place manually upgrade to ext4 by setting a couple of file system flags and running e4defrag (also converts).