I had a power failure and it took my Raq4 down. On reboot, it no longer sees the second (websites)(hdc1) drive.
I get the email "The following disks and sites are not attached and enabled:
[ST320413A-20G]".

I went into the Raq GUI and it showed unmounted. I tried to remount (enable) it in the GUI but it continued to show an error condition (it would not save), so I went in via SSH and re mounted. Then it showed in the GUI again too.

Here is my fstab file:

/dev/hda1 / ext2 defaults 1 1
/dev/hda6 swap swap defaults 0 0
/dev/hda3 /var ext2 defaults,nosuid 2 2
/dev/hda4 /home ext2 defaults,usrquota,grpquota,grpid 3 3
none /proc proc defaults 0 0
none /dev/pts devpts gid=5,mode=0622 0 0
/dev/hdc1 /vol/ST320413A-20G ext2 noauto,usrquota,grpquota,grpid
0 0

Im not very good with Linux, but assume the error is the noauto statement for hdc1. Am I correct and why did this occur? Also, how come I can now longer re-mount thru the GUI? Should I change the "noauto" to "defaults"?

Thank you very much.

I run all my 2-drive systems as RAID-1, so can't tell you the usual state of that flag in /etc/fstab. I would probably change it to defaults. It is possible that it failed fsck, and was unmounted because of that, and never re-mounted. You should look in /var/log/messages and /var/log/boot.log to look for any clues. Have you run a comprehensive fsck on it?

If you have it, you may wish to also try

/usr/sbin/smartctl -a /dev/hdc1

and/or

/usr/sbin/smartctl -t /dev/hdc1

this will give you a bunch of technical data about the drive (in the first case) and run a self-test (in the second case).

No I do not have smartctl.
So as I understand my research, I need to reboot so the hdc1 is NOT mounted and then run e2fsck. That hopefully fixes the mount problem, and then I remount it. Is this correct, or do I also need to mod the fstab file?
Thanks

You can reboot or just unmount the filesystem with umount. If you want to you can download, compile, and install the smart disk tools (just google for them).

the smart disk tools are not supported by your kernel.

Yes is sounds like you need to unmount the drive and run fsck -i /dev/hdc1 and say yes to everyting...

ext2 does this sometimes.

the smart disk tools are not supported by your kernel.
The smartmontools (http://smartmontools.sourceforge.net/) do in fact work on all versions of Linux supported by the Raq. I've compiled and run them -- they work.

The smartmontools (http://smartmontools.sourceforge.net/) do in fact work on all versions of Linux supported by the Raq. I've compiled and run them -- they work.

ok so if you didn't have them and you built them (on a raq4) why did you tell him to use someting you knew he didn't have?

ok so if you didn't have them and you built them (on a raq4) why did you tell him to use someting you knew he didn't have?

Because they're easily obtainable.
Now why do you people still use RAQs again?

Because they're easily obtainable.
Now why do you people still use RAQs again?

Hi David,

While it's possible they were easily obtainable a while back as in "download and build them", at the moment it dosen't even begin to build.

+ make
make all-recursive
make[1]: Entering directory `/home/redhat/BUILD/smartmontools-5.37'
Making all in .
make[2]: Entering directory `/home/redhat/BUILD/smartmontools-5.37'
source='smartd.cpp' object='smartd.o' libtool=no \
DEPDIR=.deps depmode=gcc /bin/sh ./depcomp \
g++ -DHAVE_CONFIG_H -I. -I. -I. -DSMARTMONTOOLS_SYSCONFDIR=\"/etc\" -O2 -m486 -fno-strength-reduce -Wall -W -c -o smartd.o smartd.cpp
smartd.cpp: In function `void PrintCVS()':
smartd.cpp:313: initialization to `char *' from `const char *' discards qualifiers
make[2]: *** [smartd.o] Error 1
make[2]: Leaving directory `/home/redhat/BUILD/smartmontools-5.37'
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory `/home/redhat/BUILD/smartmontools-5.37'
make: *** [all] Error 2

I sent a e-mail about it, so hopefully we can get it fixed for gcc 2.95 (again)

reguardless... it's not going to tell him to fsck the partition.

as far as why we use em.. I guess because we own them and they still work good. But that's another thread...

Hmmm. Guess you have a broken Raq environment. Works fine on mine.

Now why do you people still use RAQs again?

The hardware is simple and reliable (if slow). The GUI is simple and easy for simple hosting. "Use it up or wear it out -- make it do, or do without." Not many 1u systems that you can get on eBay for $25 will host 50+ simple vsites.

Hmmm. Guess you have a broken Raq environment. Works fine on mine.

I thought you only ran strongbolt now?

I thought you only ran strongbolt now?
I run Raq4, Raq550, StrongBolt, RackStar, and plain CentOS+BlueQuartz on various of over a dozen machines.

But of course, CentOS+BQ (whether in the StrongBolt distro or not) is the best Raq software environment around as of now -- it has the most modern software, the best security, the best compatibility, and the most active support community.

I have run the RaQ4 cause its easy (?) cheap and bullet proof(?). I and my associates wanted simple and complete web server platforms do do simple tests and page serves - nothing fancy ...and it grew. (maybe a message there for other small ISPs).

Now that its up OK, on the assumption that ext2 will umount randomly, is there any value in changing the fstab file? It reads :
-----------------------
[root /etc]# more fstab
/dev/hda1 / ext2 defaults 1 1
/dev/hda6 swap swap defaults 0 0
/dev/hda3 /var ext2 defaults,nosuid 2 2
/dev/hda4 /home ext2 defaults,usrquota,grpquota,grpid 3 3
none /proc proc defaults 0 0
none /dev/pts devpts gid=5,mode=0622 0 0
/dev/hdc1 /vol/ST320413A-20G ext2 noauto,usrquota,grpquota,grpid
0 0
-----------

should I make it auto or default instead of noauto? or does that not have any effect?
Thank you

This is the standard fstab

[root /root]# cat /etc/fstab
/dev/hda1 / ext2 defaults 1 1
/dev/hda6 swap swap defaults 0 0
/dev/hda3 /var ext2 defaults,nosuid 2 2
/dev/hda4 /home ext2 defaults,usrquota,grpquota,grpid 3 3
none /proc proc defaults 0 0
none /dev/pts devpts gid=5,mode=0622 0 0

With that your adding
/dev/hdc1 /vol/ST320413A-20G ext2 noauto,usrquota,grpquota,grpid
0 0

and I would say you should fsck it on boot and the options ``noauto'' (do not mount when "mount -a" is given, e.g., at boot time) is not working out for you. so yes, auto would be better for mounting at boot/reboots... which you might be doing without knowing it. (the watchdog timer will do that)

/dev/hdc1 /vol/ST320413A-20G ext2 auto,usrquota,grpquota,grpid
4 4

do you really need the quota and group stuff?
/dev/hdc1 /vol/ST320413A-20G ext2 auto 4 4

excuse my lack of knowledge...I did not knowingly put in groups or guotas. Where do I find out what they are for/mean and do?

So you are saying to change the last line: /dev/hdc1 /vol/ST320413A-20G ext2 noauto,usrquota,grpquota,grpid 0 0
to:
/dev/hdc1 /vol/ST320413A-20G ext2 auto
thanks

excuse my lack of knowledge...I did not knowingly put in groups or guotas. Where do I find out what they are for/mean and do?

You just need to...
man mount


So you are saying to change the last line: /dev/hdc1 /vol/ST320413A-20G ext2 noauto,usrquota,grpquota,grpid 0 0
to:
/dev/hdc1 /vol/ST320413A-20G ext2 auto
thanks

I'm adding the numbers to the end too...
/dev/hdc1 /vol/ST320413A-20G ext2 auto 4 4