/sbin/runscript.sh : line 27 :/var/.../softlevel no such...

RCMN

during an emerge -Du world , my system start becoming very very slow.After a while all application were kill because the system was out of memory.

i had no other choice than rebooting.

but this message come up during the boot sequence

/sbin/runscript.sh : line 27 :/var/lib/init.d/softlevel no such file or directory.

so i decide to boot with the cd and understand.I found an init_old.d and replace my actual init.d i was able to reboot .I thought i was out of trouble.When i decide to finish my 'emerge -Du world'.

And the problem append again .The problem now i don't have the original init_old.d so i'm stuck here ;without softlevel in my /var/.../init.d/ and a runscript.sh who overflow my memory.

<hummm HELP!!!

RCMN · Posted: Tue Dec 09, 2003 7:41 pm Post subject:

help i don't want reinstall, after 6 month of happyness

.

what should i do ???

undertoe · n00b Joined: 27 Jan 2004 Posts: 2 Location: united states

99% of the time this should fix your problem

echo default > /var/lib/init.d/softlevel

stalcair · n00b Joined: 07 Aug 2003 Posts: 57

I have been playing around with this for 2 days. Perhaps I can clone myself and one does Googling, the other searches here, the other plays around with various settings, and the last tries different kernel configurations.

I didn't develop this problem on my home computer (an a7n8x) but on an eServer xSeries 305 I get a rather nasty variant of what is referred to (yet strangly never really fixed) as "the depscan problem"

Note that this may be another issue that is causing the depscan issue, but I am at a loss now.

My system and setup attempts:
lvm2
love-sources (multiple ones in the 2.6.5 series)
UDEV, no devfs (using the gentoo=nodevfs kernel option from grub)
2.4.22 (or what is the current stable version) AND 2.6.4 linux-headers
stable (1.8.6.13) and unstable (1.8.6.13-r1)
manually created /var/lib/init.d/softlevel (with default)
verified that my /etc/rc.conf does NOT have uncommented PROTOCOL lines
/boot, swap and / are all regular partitions
device mapper is compiled in, not a module

I just started building this a few days ago, and have had trouble since I booted it the first time (post LiveCD)

What I have now is a system that does not mount the logical partitions at boot, thus all sorts of nastiness because files can't be found (the /usr, /home, /opt, /var, and /tmp are all lvm2)

I can't even start up sshd, even after the manual creation of /var/lib/init.d/softlevel with "default" and /sbin/depscan.sh is run

If I do run /etc/init.d/sshd after these two changes are made, then I end up with my net.lo re-run (yet it is already there and set to load at boot, per rc-update -s) Also it claims that sshd is started, but yet it is not. net.lo and sshd are then both present (and only them) in /var/lib/init.d/started/

When booting, the ncurses output claims that eth0 is not setup (dhcp) yet it works fine

Other significant rc scripts are as follows:
hotplug boot
net.eth0 boot

Another interesting tidbit is that when I try to start syslog-ng it is as if I have just then spawned a chane in runlevel, my "remaining partitions" are mounted as read-only and such yet that is not really happening that I can tell.

I appologize for no actual logs or dumps, but the machine is currently down and I am running the livecd to try and do everything from a chroot, thus giving me ssh access.

I may have just missed a step in installing either UDEV, LVM2, the 2.6 kernel or anything else but since I have been installing this setup on 3 other computers recently with no problems I am doubting it... then again I am severely sleep deprived and so anything is possible.

I am hoping that if this magically solves itself that someone can help me figure out WHY. One of the things I love about Gentoo is how much "easier" it is to understand the guts of your system... yet here I really would first like to get back to work and then try to figure out the details... that sound odd enough?

Thanks for any information

stalcair · n00b Joined: 07 Aug 2003 Posts: 57

*tap* *tap*
Hello? Is this thing on?

I've got net access now... so what logs, dumps, etc would be useful here?

As the subject says, if you have had this problem please post what your solution was. If nothing else perhaps there can be a less fragmented area for others with this problem to search for. (especially considering the poor search capability phpbb provides)

I have searched for a very long time and come up with no actual solution. Read: what the cause was, how to fix it (not work around), and avoid it in the future.

search terms used:
depscan
depscan.sh
error
runscript
runscript.sh
softlevel
lvm
lvm2
love sources
love
2.6 (haha, thought I would torture myself with that one)
2004
livecd

and basically combinations of those

Please help me find the solution to this problem, my curiosity is as strong as my need to get back to work on that box.

If the answer is to rebuild, then I will put Suse on now and try later. Gentoo is not the 20 minute setup distro, but I have been working on this for many days now and perhaps I should stop being stubborn?

Again, please post your knowledge here... the forums are pointless if people do not put what is in their head onto them.

invis · n00b Joined: 05 Jul 2003 Posts: 7

I had the same problem and what worked for me was updating hotplug to hotplug-20040105

... unfortunately I still get errors with hotplug segfaulting.. not sure why.

Decibels · Veteran Joined: 16 Aug 2002 Posts: 1623 Location: U.S.A.

I get that same error as of recently:

/sbin/runscript.sh : line 27 :/var/lib/init.d/softlevel no such file or directory.

Looks like a timing issue, because boots up fine and everything works. The file is also there and has default in it. Running pure udev system, but no lvm. Guessing that this file is created during boot.
_________________
Support bacteria – they’re the only culture some people have.”

– Steven Wright

snowpatch · n00b Joined: 09 Mar 2003 Posts: 38 Location: Philadelphia

I had the same softlevel error along with depcache errors after upgrading to baselayout-1.8.8 and hotplug-20040401. I tried downgrading baselayout, but the thing that worked was downgrading hotplug to 20040329.

Decibels · Veteran Joined: 16 Aug 2002 Posts: 1623 Location: U.S.A.

From other post have seen, it is beginning to look like the same hotplug you have. I have an earlier version of baselayout 1.8.6.13-r1 and had it awhile.
Did recently get the 20040401 hotplug.

Also put a 'ls -l /var/lib/init.d >> test.txt' in runscript.sh and from the very first one 'softlevel' is there. When have time going to look into it a little further.
_________________
Support bacteria – they’re the only culture some people have.”

– Steven Wright

Nate_S · Guru Joined: 18 Mar 2004 Posts: 414

I had this problem about a month ago, but in my case it was not the fault of the os, but rather it just so happened that my hard drive had a block fail right at that location. reiserfsck --rebuild-tree made it work like normal again

dopey · Apprentice Joined: 10 Feb 2003 Posts: 235

I ran into the same problem last night. Poked around and found out what's causing it. It's caused by hotplug-20040401 I think. It didn't seem to happen with the older version of hotplug and I'm not sure why. However, here's what's causing it for me:
1) I autoload my network modules in /etc/modules.autoload.d/kernel-2.6
2) during boot, these modules are loaded. Due to hotplug support in the kernel, this triggers events when the network devices are added. Once this happens, hotplug calls net.eth0 and net.eth1 (I have two net devices).

Now normally, this isn't a problem, but I have a seperate partition for my /var filesystem. During the boot process, local filesystems get mounted AFTER modules are loaded. The problem occurs because the net scripts are run before local fllesystems are mounted. As such /var/lib/init.d is empty because normally the stuff is populated after /var is mounted. This results in the goofy messages happening. I just ignore them, since it doesn't matter because the net scripts are run as part of the default runlevel anyways.

Hope this makes sense. It might explain what you're running into.

snowpatch · n00b Joined: 09 Mar 2003 Posts: 38 Location: Philadelphia

As far as it being a hotplug issue, I found this bug:
https://bugs.gentoo.org/show_bug.cgi?id=47111 that describes the same sort of thing.

Like Dopey, I have my network card module loaded at boot. However, the only partitions I have are /boot(not mounted), / and /home.

But what is different between hotplug-20040329 and 20040401?

stalcair · n00b Joined: 07 Aug 2003 Posts: 57

I had a chance to revisit this recently. I reinstalled the system (making about 3 times) but this was before the 1.8.10 baselayout came out and I believe before any changes to hotplug as well.

First, let me say that I tried many combinations of hotplug and baselayout versions with never a good result.

What I found was to perform this action, thanks to Decibels.

dopey · Apprentice Joined: 10 Feb 2003 Posts: 235

dswhite42 · n00b Joined: 20 Dec 2003 Posts: 7

dopey · Apprentice Joined: 10 Feb 2003 Posts: 235

sortkatt · n00b Joined: 05 Mar 2003 Posts: 13

snowpatch · n00b Joined: 09 Mar 2003 Posts: 38 Location: Philadelphia

I have the newest baselayout-1.8.11 but it did not fix my problems either. I get errors about /var/init.d/softlevel not there and also errors about /var/lib/ being a read-only filesystem. This is all with hotplug-20040401 and 2.6.4-ck2 with pure udev. Any ideas? Should I just wait for a fix in hotplug?

sortkatt · n00b Joined: 05 Mar 2003 Posts: 13

Hey! Something wonderful happened. The new udev-package is out, and it fixed all my issues. (udev025 I think) Just merge new version and all will be fine. Blissfull ignorance worked out in the end :oops:

Been reading up on lvm, seems like a neat idea... Will try to set it up another time.

Emerge -pul udev shows that the nevest version of udev is marked stable

I would call that quite a fast switch

UPDATE: No it isn't fixed. Just a fluke boot where everything was timed right

*waiting for update of hotplug like the rest*
_________________
-this is a signature

leo.fontenelle

kallamej · Posted: Thu Jun 16, 2005 9:23 pm Post subject:

There is also an FAQ for this: https://forums.gentoo.org/viewtopic-t-53821.html

Moved from Portage & Programming to Duplicate Threads.
_________________
Please read our FAQ Forum, it answers many of your questions.
irc: #gentoo-forums on irc.libera.chat