| View previous topic :: View next topic |
| Author |
Message |
devzero_DE
n00b
Joined: 06 Jun 2004
Posts: 36
Location: Cologne / Germany
|
 Posted: Thu Jun 29, 2006 10:01 am Post subject: Software RAID - strange bootup message? |
 |
|
Hey,
I dont know if this is the right place for my post but I hope so.
I created two raid1 and one raid5 with mdadm.
| Code: | dev0 ~ # cat /etc/mdadm.conf
ARRAY /dev/md3 level=raid1 num-devices=2 UUID=d01af3da:4fd73b7c:540e0b0b:116c5ecf
ARRAY /dev/md1 level=raid1 num-devices=2 UUID=d943efe5:95e14857:70a20340:82cebf07
ARRAY /dev/md5 level=raid5 num-devices=3 UUID=3566b48e:4634ba25:1febb160:dc90a47c |
The arrays are up and working after reboot - so theres no problem.
No problem? There is something strange on bootup..
| Code: | * Starting up RAID devices (mdadm) ... [!!]
mdadm: No arrays found in config file |
What config file? At this point, the Arrays are already up. |
|
| Back to top |
|
 |
poncio
n00b
Joined: 02 Nov 2004
Posts: 19
|
| Posted: Thu Jun 29, 2006 1:10 pm Post subject: |
|
|
A look at the /etc/init.d/mdadm script suggests that it is looking for a config file (maybe in /etc/conf.d/) but I could not find it in my install.
Having said that, I also run software raid without the mdadm script with no problems. |
|
| Back to top |
|
|
devzero_DE
n00b
Joined: 06 Jun 2004
Posts: 36
Location: Cologne / Germany
|
| Posted: Thu Jun 29, 2006 1:40 pm Post subject: |
|
|
The mdadm init script isn't started and isn't linked to default or boot.
I think the mdadm package is only used to manage the raids or for stauts mails.
I don't know which scripts prints this error. |
|
| Back to top |
|
|
poncio
n00b
Joined: 02 Nov 2004
Posts: 19
|
| Posted: Thu Jun 29, 2006 2:51 pm Post subject: |
|
|
No script that I could find generates that error.
The mdadm command is the only who can output that error message.
Could you post at what point in the boot process you get the error? |
|
| Back to top |
|
|
piwacet
Guru
Joined: 30 Dec 2004
Posts: 486
|
| Posted: Sat Jul 08, 2006 7:04 pm Post subject: |
|
|
Hi. I have this problem too. I have 2 sata disks, with RAID1 on /md1 (sda1 and sdb1), swap on the next 2 partitions, and then / on md3 (sda3 and sdb3). Since upgrading to the mdadm-2.5* series, I get this error at boot, at this time:
Finalizing UDEV configuration
Mounting devpts @ /dev/pts...
Starting up RAID devices...
mdadm: no arrays found in config file
checking root filesystem
My /etc/mdadm.conf has:
| Code: | ARRAY /dev/md3 level=raid0 num-devices=2 UUID=485d0c26:ff42708b:d5d7cc49:33228f77
ARRAY /dev/md1 level=raid1 num-devices=2 UUID=d22154e5:10c12275:3f5723c9:5c82df05 |
As far as I can tell things still work correctly.
Any thoughts? |
|
| Back to top |
|
|
troymc
Guru
Joined: 22 Mar 2006
Posts: 553
|
| Posted: Sat Jul 08, 2006 8:54 pm Post subject: |
|
|
This is one of those niggling little messages that look bad but really mean nothing. 
Basically mdadm is scanning your config file, and cannot find any arrays that have not already been started.
Why? Because the kernel code has already started all the arrays at boot.
You can duplicate this error by running mdadm --assemble --scan at the command line.
| Code: |
# cat /proc/mdstat
Personalities : [linear] [raid0] [raid1]
md1 : active raid0 sdb4[1] sda4[0]
97707136 blocks 64k chunks
md0 : active raid1 sdb1[1] sda1[0]
136448 blocks [2/2] [UU]
bitmap: 0/17 pages [0KB], 4KB chunk
unused devices: <none>
#
# cat /etc/mdadm.conf
DEVICE partitions
ARRAY /dev/md0 devices=/dev/sda1,/dev/sdb1
ARRAY /dev/md1 devices=/dev/sda4,/dev/sdb4
MAILADDR root@localhost
#
# mdadm --assemble --scan
mdadm: No arrays found in config file
|
Don't Panic! 
My arrays are fine, they're just already running.
troymc |
|
| Back to top |
|
|
piwacet
Guru
Joined: 30 Dec 2004
Posts: 486
|
| Posted: Sat Jul 08, 2006 8:58 pm Post subject: |
|
|
| Thanks! |
|
| Back to top |
|
|
R!tman
Veteran
Joined: 18 Dec 2003
Posts: 1303
Location: Zurich, Switzerland
|
| Posted: Sun Jul 16, 2006 1:04 pm Post subject: |
|
|
Isn't there a way to at least get rid of the error message?
Really strange, I have no idea what starts this in the first place. |
|
| Back to top |
|
|
intgr
Apprentice
Joined: 23 Jun 2004
Posts: 225
Location: Earth, Sol, Milky Way
|
| Posted: Sun Jul 16, 2006 3:51 pm Post subject: |
|
|
| R!tman wrote: | | Isn't there a way to at least get rid of the error message? |
Sounds like this would do it:
| Code: | | rc-update del mdadm default |
| Quote: | | Really strange, I have no idea what starts this in the first place. |
Yeah, the kernel does RAID autodetection these days. |
|
| Back to top |
|
|
R!tman
Veteran
Joined: 18 Dec 2003
Posts: 1303
Location: Zurich, Switzerland
|
| Posted: Sun Jul 16, 2006 8:39 pm Post subject: |
|
|
| intgr wrote: | | R!tman wrote: | | Isn't there a way to at least get rid of the error message? |
Sounds like this would do it:
| Code: | | rc-update del mdadm default |
|
No, that's not it. mdadm is not started during boot! Feels strange now too? |
|
| Back to top |
|
|
DNAspark99
Guru
Joined: 03 Sep 2004
Posts: 321
|
| Posted: Tue Jul 18, 2006 4:57 pm Post subject: |
|
|
I have the same message displayed at boot - still no fix?
also, although all arrays seem fine:
| Code: |
Personalities : [raid1]
md1 : active raid1 sdb1[1] sda1[0]
56128 blocks [2/2] [UU]
md2 : active raid1 sdb2[1] sda2[0]
1003968 blocks [2/2] [UU]
md3 : active raid1 sdb5[1] sda5[0]
15004608 blocks [2/2] [UU]
md4 : active raid1 sdb6[1] sda6[0]
15004608 blocks [2/2] [UU]
md5 : active raid1 sdb7[1] sda7[0]
47078336 blocks [2/2] [UU]
unused devices: <none> |
I do have this somewhat unsettling message in dmesg output:
| Code: |
md: Autodetecting RAID arrays.
md: autorun ...
md: considering sdb7 ...
md: adding sdb7 ...
md: sdb6 has different UUID to sdb7
md: sdb5 has different UUID to sdb7
md: sdb2 has different UUID to sdb7
md: sdb1 has different UUID to sdb7
md: adding sda7 ...
md: sda6 has different UUID to sdb7
md: sda5 has different UUID to sdb7
md: sda2 has different UUID to sdb7
md: sda1 has different UUID to sdb7
md: created md5
md: bind<sda7>
md: bind<sdb7>
md: running: <sdb7><sda7>
raid1: raid set md5 active with 2 out of 2 mirrors
md: considering sdb6 ...
md: adding sdb6 ...
md: sdb5 has different UUID to sdb6
md: sdb2 has different UUID to sdb6
md: sdb1 has different UUID to sdb6
md: adding sda6 ...
md: sda5 has different UUID to sdb6
md: sda2 has different UUID to sdb6
md: sda1 has different UUID to sdb6
md: created md4
md: bind<sda6>
md: bind<sdb6>
md: running: <sdb6><sda6>
raid1: raid set md4 active with 2 out of 2 mirrors
md: considering sdb5 ...
md: adding sdb5 ...
md: sdb2 has different UUID to sdb5
md: sdb1 has different UUID to sdb5
md: adding sda5 ...
md: sda2 has different UUID to sdb5
md: sda1 has different UUID to sdb5
md: created md3
md: bind<sda5>
md: bind<sdb5>
md: running: <sdb5><sda5>
raid1: raid set md3 active with 2 out of 2 mirrors
md: considering sdb2 ...
md: adding sdb2 ...
md: sdb1 has different UUID to sdb2
md: adding sda2 ...
md: sda1 has different UUID to sdb2
md: created md2
md: bind<sda2>
md: bind<sdb2>
md: running: <sdb2><sda2>
raid1: raid set md2 active with 2 out of 2 mirrors
md: considering sdb1 ...
md: adding sdb1 ...
md: adding sda1 ...
md: created md1
md: bind<sda1>
md: bind<sdb1>
md: running: <sdb1><sda1>
raid1: raid set md1 active with 2 out of 2 mirrors
md: ... autorun DONE.
md: Loading md3: /dev/sda5
md: couldn't update array info. -22
md: could not bd_claim sda5.
md: md_import_device returned -16
md: could not bd_claim sdb5.
md: md_import_device returned -16
md: starting md3 failed |
md3 is my / (root) partition... what's with this error, failure to start md3? Is this just a convoluted way of saying it's 'allready started' ? |
|
| Back to top |
|
|
intgr
Apprentice
Joined: 23 Jun 2004
Posts: 225
Location: Earth, Sol, Milky Way
|
| Posted: Tue Jul 18, 2006 5:19 pm Post subject: |
|
|
| DNAspark99 wrote: | | Code: | md: Loading md3: /dev/sda5
md: couldn't update array info. -22
md: could not bd_claim sda5.
md: md_import_device returned -16
md: could not bd_claim sdb5.
md: md_import_device returned -16
md: starting md3 failed |
md3 is my / (root) partition... what's with this error, failure to start md3? Is this just a convoluted way of saying it's 'allready started' ? |
errno 22 is EINVAL (invalid argument) which usually indicates something dodgy in the (kernel) code. Looking at the source, it doesn't look like anything critical, but there are a lot of paths that can end up with this error.
The other error, 16 is EBUSY which indicates (suprisingly) that the device is busy. If it's your root partition, have you specified root=/dev/sda5 instead of root=/dev/md3 by any chance? This could be the source for both warnings.
It's odd that md3 actually starts after those messages.
Edit: A Google search indicates that these messages will appear if you have "md=3,/dev/sda5,/dev/sdb5" in your kernel arguments with embedded-superblock md partitions - the RAID will be started once from those parameters and subsequently by the md autodetection (which fails). |
|
| Back to top |
|
|
DNAspark99
Guru
Joined: 03 Sep 2004
Posts: 321
|
| Posted: Tue Jul 18, 2006 6:45 pm Post subject: |
|
|
Aaah, yes, my grub.conf contains:
| Code: | | kernel /vmlinuz root=/dev/md3 md=3,/dev/sda5,/dev/sdb5 |
Trimming out the md=3,... removes the errors as well. Thanks. |
|
| Back to top |
|
|
pota
n00b
Joined: 25 Jan 2005
Posts: 22
|
| Posted: Fri Jul 28, 2006 2:42 pm Post subject: |
|
|
| Code: | * Starting up RAID devices (mdadm) ... [!!]
mdadm: No arrays found in config file |
this message is generated by raid_start.sh, if you don't want this just remove a 'ARRAY /dev/mdX...' line from mdadm.conf, be warned however that without configuration you can't start your RAID with 'mdadm -As /dev/mdX'... |
|
| Back to top |
|
|
96140
Retired Dev
Joined: 23 Jan 2005
Posts: 1324
|
| Posted: Wed Oct 25, 2006 2:52 am Post subject: |
|
|
| Thanks, I just got this weird message myself, despite the fact that my RAID1 array is obviously working; checking /proc/mdstat confirms that everything is healthy. We need a smarter relationship between the kernel's autodetection and the rc script, methinks. |
|
| Back to top |
|
|
neysx
Retired Dev
Joined: 27 Jan 2003
Posts: 795
|
| Posted: Wed Oct 25, 2006 11:59 am Post subject: |
|
|
I guess you could remove raid from RC_VOLUME_ORDER in /etc/conf.d/rc. I haven't tried and I'm not sure what problems could occur as a result of not calling mdadm -S to stop your raid when shutting down. / cannot be shut down anyway and I've never had any trouble.
Another way is to edit /lib/rcscripts/addons/raid-start.sh to make it call mdadm -As only when some devices defined in /etc/mdadm.conf are not active yet | Code: | # Start software raid with mdadm (new school)
if [[ -x /sbin/mdadm && -f /etc/mdadm.conf ]] ; then
devs=$(awk '/^md[[:digit:]]+ : active/{activ["/dev/" $1]=1} /^[[:space:]]*ARRAY/ {if (!activ[$2])print $2 }' /proc/mdstat /etc/mdadm.conf)
if [[ -n ${devs} ]] ; then
create_devs ${devs}
ebegin "Starting up RAID devices (mdadm)"
output=$(mdadm -As 2>&1)
ret=$?
[[ ${ret} -ne 0 ]] && echo "${output}"
eend ${ret}
fi
fi |
The only change is on the awk line
Hth |
|
| Back to top |
|
|
96140
Retired Dev
Joined: 23 Jan 2005
Posts: 1324
|
| Posted: Wed Oct 25, 2006 7:04 pm Post subject: |
|
|
| neysx wrote: | | I guess you could remove raid from RC_VOLUME_ORDER in /etc/conf.d/rc. I haven't tried and I'm not sure what problems could occur as a result of not calling mdadm -S to stop your raid when shutting down. / cannot be shut down anyway and I've never had any trouble. |
Removing it from VOLUME_ORDER sounds a little edgy; I'd prefer to let someone else try it out before me. I already get some interesting messages about not being able to stop /dev/md1 (/boot) on shutdown, whether or not it's mounted. Of course, then a second later it tells me everything is fine & shut down, and then shutdown continues as normal. Again, there's something weird in the kernel auto-RAID stuff vs. the init script-provided RAID.
Good to know that md2 (/) can't be shut down anyway; makes sense. If I ever see a related message, now I'll know why.  |
|
| Back to top |
|
|
Aysen
Apprentice
Joined: 18 May 2005
Posts: 187
Location: Poland
|
| Posted: Sat Oct 28, 2006 6:45 pm Post subject: |
|
|
| neysx wrote: | | / cannot be shut down anyway and I've never had any trouble. |
Oh, so is this why I always get a | Code: | | mdadm: failed to stop array /dev/md3 Device or resource busy |
error on a reboot/halt, yet my array is perfectly healthy (at least I believe so  )?
I'm also getting that mdadm error during boot, good to know it's nothing wrong.
This thread's been very informative for me, thank you. |
|
| Back to top |
|
|
tgh
Apprentice
Joined: 05 Oct 2005
Posts: 222
|
| Posted: Thu Nov 02, 2006 9:44 pm Post subject: |
|
|
mdadm does not require a config file if you use the newer RAID types. It puts a special RAID ID on each drive so that it knows which drive goes with which. However, if you really want to create a config file...
| Code: | | mdadm --detail --scan >> /etc/mdadm.conf |
|
|
| Back to top |
|
|
Aysen
Apprentice
Joined: 18 May 2005
Posts: 187
Location: Poland
|
| Posted: Fri Nov 10, 2006 7:27 pm Post subject: |
|
|
| tgh wrote: | mdadm does not require a config file if you use the newer RAID types. It puts a special RAID ID on each drive so that it knows which drive goes with which. However, if you really want to create a config file...
| Code: | | mdadm --detail --scan >> /etc/mdadm.conf |
|
I knew I didn't actually need that config file, but the HOWTO I followed to set up mdadm for the first time instructed to create it (exactly the way you wrote), so I just did it. Now, after removing that file, I don't get those errors anymore.
Thank you! |
|
| Back to top |
|
|
96140
Retired Dev
Joined: 23 Jan 2005
Posts: 1324
|
| Posted: Fri Nov 10, 2006 9:32 pm Post subject: |
|
|
| Note to all: you can live without the config file and/or init script if and only if you have set your RAID partition type to fd (Linux RAID autodetect); this is the only way you can get the kernel to do its autodetect/setup magic. |
|
| Back to top |
|
|
overkll
Veteran
Joined: 21 Sep 2004
Posts: 1249
Location: Austin, Texas
|
| Posted: Sat Dec 02, 2006 9:09 pm Post subject: |
|
|
| Aysen wrote: | | tgh wrote: | mdadm does not require a config file if you use the newer RAID types. It puts a special RAID ID on each drive so that it knows which drive goes with which. However, if you really want to create a config file...
| Code: | | mdadm --detail --scan >> /etc/mdadm.conf |
|
I knew I didn't actually need that config file, but the HOWTO I followed to set up mdadm for the first time instructed to create it (exactly the way you wrote), so I just did it. Now, after removing that file, I don't get those errors anymore.
Thank you! |
Works here too. Only one problem - the MAILADDR variable in /etc/mdadm.conf.
So I added
to MDADM_OPTS in /etc/conf.d/mdadm to insure I still get notification messages from the mdadm monitor daemon. |
|
| Back to top |
|
|
avieth
Veteran
Joined: 17 Sep 2004
Posts: 1945
Location: Canada
|
| Posted: Wed Dec 13, 2006 4:02 am Post subject: |
|
|
This is a little unrelated, but after starting that mdadm initscript and getting the reported warning (no arrays found) the init process proceeds to check my filesystems.
EVERY single time I boot it outputs at least 10 lines of text, and pauses on 'replaying journal...' for about 5-8 seconds. This is really annoying, as I like a speedy boot. Any ideas why my reiserfs filesystem is doing this? |
|
| Back to top |
|
|
b00zy
n00b
Joined: 31 Oct 2004
Posts: 46
|
| Posted: Wed Jan 17, 2007 10:48 pm Post subject: RAID not unmounting properly |
|
|
I have a RAID setup like this...
| Code: | /dev/md1: /boot
/dev/md2: /swap
/dev/md3: / |
At the end of the shutdown sequence, /dev/md1 and /dev/md2 are unmounted. However, an error is returned when unmounting /dev/md3 "Device or resource busy."
Everything in boot and normal operation works fine other than that error. |
|
| Back to top |
|
|
b00zy
n00b
Joined: 31 Oct 2004
Posts: 46
|
| Posted: Thu Jan 18, 2007 7:59 am Post subject: |
|
|
| Aysen wrote: | | neysx wrote: | | / cannot be shut down anyway and I've never had any trouble. |
Oh, so is this why I always get a | Code: | | mdadm: failed to stop array /dev/md3 Device or resource busy |
error on a reboot/halt, yet my array is perfectly healthy (at least I believe so )?
I'm also getting that mdadm error during boot, good to know it's nothing wrong.
This thread's been very informative for me, thank you. |
Is that true? Is there absolutely no way to quell this error? |
|
| Back to top |
|
|
|
Kernel & Hardware
