Unfortunately no one can be told what fun_plug is - you have to see it for yourself.
You are not logged in.
Here's an update. I fine the log at the root of volume2, not where I was expecting. The last few lines were:
Filesystems ready to check: /dev/sdb2
Start checking /dev/sdb2 at 19:02:02
/dev/sdb2: Adding dirhash hint to filesystem.
/dev/sdb2 contains a file system with errors, check forced.
/dev/sdb2: Inode 8769 is in use, but has dtime set. FIXED.
/dev/sdb2: Inode 8769 has imagic flag set.
/dev/sdb2: UNEXPECTED INCONSISTENCY; RUN fsck MANUALLY.
(i.e., without -a or -p options)
e2fsck on /dev/sdb2 status: 4, PARTITION /dev/sdb2 CHECK ERROR, USER ATTENTION NEEDED!
PARTITION /dev/sdb2 MOUNTED READ ONLY
Done, filesystems mounted and services restarted at Tue Jun 2 19:03:29 GMT 2009.
------------
OK, so I go to run the script manually, and the drive lights flash orange, but I'm locked out of telnet. Putty just keeps quitting while trying to log on.
How do I get manual mode to work?
Pete
Offline
pscott wrote:
I just tried this yesterday, and it ran much quicker than I expected. I have 2 1TB drives in my box, each as an individual volume.
Much quicker? hmmm, if your filesystem needs checking if should take about an hour for each disk! That's why dlink/conceptronic removed that needed feature from their firmware, "availability". (and because ext3 was not working in the initial versions)
With ext3, fsck "its not need", only every xxx mounts or yyy months, it depends on you, check current values with "tune2fs -l /dev/sda2", look at Check interval: and Maximum mount count.
However, now my entire 2nd volume is marked as read-only. Checking with the user privileges page, the drive is marked as R/W. What gives, and how do I return it to R/W?
If it is marked as read-only is because fsck failed on it, and yafsck mounted it that way to avoid further errors. Have you read the produced FSCK.log file at /mnt/HD_a2? What does it say? In certain circunstances it is not possible to write the file, as when the filesystem is remounted read-only.
The "mount" command should give more details than the web page.
You can try to do a clean reboot... or diagnose first the problem, There are commands to force a remount in read/write mode, but you must know what you are doing before issuing them. That's why I don't quote them here (for now). ;-)
Offline
Here's what the end of the fsck.log at /mnt/HD_a2 said:
-----------------------------------------
Filesystems ready to check: /dev/sdb2
Start checking /dev/sdb2 at 19:02:02
/dev/sdb2: Adding dirhash hint to filesystem.
/dev/sdb2 contains a file system with errors, check forced.
/dev/sdb2: Inode 8769 is in use, but has dtime set. FIXED.
/dev/sdb2: Inode 8769 has imagic flag set.
/dev/sdb2: UNEXPECTED INCONSISTENCY; RUN fsck MANUALLY.
(i.e., without -a or -p options)
e2fsck on /dev/sdb2 status: 4, PARTITION /dev/sdb2 CHECK ERROR, USER ATTENTION NEEDED!
PARTITION /dev/sdb2 MOUNTED READ ONLY
Done, filesystems mounted and services restarted at Tue Jun 2 19:03:29 GMT 2009.
-----------------------------------------
This is what I posted in the last message.
My problem is that the manual mode does not work. My telnet session just hangs with no response after invoking the manual mode. I've got ffp 0.5, and your script seems to be in the correct folder. The first few lines of your script are echoed, then telenet session is suspended, as expected.
After that, drives lights go orange, but I cannot telnet in again, as there is no response when starting a new telnet session.
I've rebooted, the drive is mounted normally, but still has errors. The results from the tune2fs command is:
Filesystem volume name: <none>
Last mounted on: <not available>
Filesystem UUID: ab4b454c-0563-4070-922e-dbb5a4d8926e
Filesystem magic number: 0xEF53
Filesystem revision #: 1 (dynamic)
Filesystem features: dir_index filetype sparse_super large_file
Default mount options: (none)
Filesystem state: not clean with errors
Errors behavior: Continue
Filesystem OS type: Linux
Inode count: 122044416
Block count: 244057471
Reserved block count: 0
Free blocks: 4513000
Free inodes: 121831885
First block: 0
Block size: 4096
Fragment size: 4096
Blocks per group: 32768
Fragments per group: 32768
Inodes per group: 16384
Inode blocks per group: 512
Filesystem created: Mon Feb 16 11:57:06 2009
Last mount time: Tue Feb 17 20:49:27 2009
Last write time: Wed Jun 3 20:02:56 2009
Mount count: 118
Maximum mount count: 39
Last checked: Mon Feb 16 11:57:06 2009
Check interval: 15552000 (6 months)
Next check after: Sat Aug 15 11:57:06 2009
Reserved blocks uid: 0 (user root)
Reserved blocks gid: 0 (group root)
First inode: 11
Inode size: 128
Default directory hash: tea
Directory Hash Seed: d9275d2c-8b0d-4a9a-b290-eb15df5f3c16
My question is though: why can I not get the manual mode to work? Obviously the script is finding errors on the second drive, which is why it's mounted read-only.
Offline
pscott wrote:
Here's what the end of the fsck.log at /mnt/HD_a2 said:
-----------------------------------------
Filesystems ready to check: /dev/sdb2
Start checking /dev/sdb2 at 19:02:02
/dev/sdb2: Adding dirhash hint to filesystem.
/dev/sdb2 contains a file system with errors, check forced.
/dev/sdb2: Inode 8769 is in use, but has dtime set. FIXED.
/dev/sdb2: Inode 8769 has imagic flag set.
/dev/sdb2: UNEXPECTED INCONSISTENCY; RUN fsck MANUALLY.
(i.e., without -a or -p options)
e2fsck on /dev/sdb2 status: 4, PARTITION /dev/sdb2 CHECK ERROR, USER ATTENTION NEEDED!
PARTITION /dev/sdb2 MOUNTED READ ONLY
Done, filesystems mounted and services restarted at Tue Jun 2 19:03:29 GMT 2009.
-----------------------------------------
So you have a real problem at hands.
I would recommend to backup your data while you still can access it.
The problem with running fsck in manual mode, as itself recomends (these are fsck messages, not my script):
/dev/sdb2: UNEXPECTED INCONSISTENCY; RUN fsck MANUALLY.
(i.e., without -a or -p options)
is that it is going to make you some really strange questions, and you don't know what you to answer. Most people don't, so fsck has an option, "-y", to always give (on behalf of you) an YES answer to all questions. Of course it can be the wrong answer... There is also a "-n" option.
The "good" news is that the faulty disk is the second one, so you can run fsck without special needs.
First kill all processes, except your telnet session
Then "cd /"
Then "umount /dev/sdb2"
if it says its busy, there is still some process accessing it, kill it.
Then "fsck /dev/sdb2" and the strange questions start appearing. It can be just one, two or three or one-thousand...
At the end fsck will almost certainty tell you to reboot. Do it, don't try mounting the partition. And that's it.
I should refrain myself from given advices in this matter. In my experience, "fsck -n /dev/sdb2" doesn't solve any problem, while "fsck -y /dev/sdb2" produces a lot of frightening messages. Its your choice. The other is do nothing. Up to you.
My problem is that the manual mode does not work. My telnet session just hangs with no response after invoking the manual mode. I've got ffp 0.5, and your script seems to be in the correct folder. The first few lines of your script are echoed, then telenet session is suspended, as expected.
After that, drives lights go orange, but I cannot telnet in again, as there is no response when starting a new telnet session.
...
My question is though: why can I not get the manual mode to work? Obviously the script is finding errors on the second drive, which is why it's mounted read-only.
We can try to diagnose that latter.
Last edited by jcard (2009-06-04 10:46:45)
Offline
Thanks. All my data is backed up, so I will try your suggestion today.
Pete
Offline