wengi
Full Member
Posts: 102
|
Post by wengi on Jul 7, 2008 15:42:55 GMT 7
There are processes running on /raidbuechse:~# ps axf | grep -i raid
2110 root SW [md1_raid1]
2112 root SW [md0_raid1]
2161 root 608 S /bin/sh /app/bin/raid_m2
2169 root 556 S /bin/sh /img/bin/raid_full
2854 root 700 S /raid/module/NFSD/sbin/rpc.mountd
6095 root 460 S grep -i raid
but even stopping / killing them did not help here.
wengi
|
|
|
|
Post by lewinb on Jul 8, 2008 5:59:13 GMT 7
Are these running even if I have all services turned off in the web interface? And can you tell what these processes are?
It is a little odd to me that the system is designed to be loaded entirely from flash ram, yet it's running a process located on /raid/module/
... Does that mean that if both disks are replaced, the modules have to be reinstalled as well?
|
|
wengi
Full Member
Posts: 102
|
Post by wengi on Jul 8, 2008 14:31:55 GMT 7
Hi,
/raid/module/NFSD/sbin/rpc.mountd is the NFS module. No Prob to stop/disable this if it is installed. So this one will not run if disabled in webinterface.
You will not have to reinstall the modules if we can resize container/partition.
You will have to reinstall the modules if you use new HDDs and copy the data back from the old HDDs.
I have no idea about raid_m2 and raid_full. This seems to be thecus specific.
Maybe one of thr pro's of the board can tell us something about md1 and the both above.
wengi
|
|
|
|
Post by dbridges on Jul 28, 2008 13:16:02 GMT 7
wow... it's like these boards have been abandoned! dbridges, where are you, oh wise one?! :-)  Sorry. But this one is beyond me anyway. All I know is what i've learn't while playing with my own n2100. |
|
|
|
Post by lewinb on Aug 12, 2008 9:43:03 GMT 7
Well, just out of curiosity, let's look at a different possibility...
The raid will operate in degraded mode with a drive missing. I think I remember reading that the n2100 is a software raid. That being the case, is it possible that the drive could be mounted via a usb enclosure by another OS (Mac/Ubuntu, etc.) which has support for ext2 or ext3?
The Drive Utility with Mac os is capable of creating raids and altering partitions, so could it mount a raid created on another linux box? And then increase the partition?
I don't know how much time/effort this would save... I guess it would resize the partition without having to copy/recopy, then rebuild. So that would be a start.
|
|
|
|
Post by wookienz on Apr 12, 2009 18:11:08 GMT 7
I have a 4100+
I found this page via google thus my posting here.
What i did for my raid 5 which sounds like we have a similar problem, is i changed all the drives out - 500gbs for 250gbs. However the raid was still at 750gb (3x250, 1 for parity). So to increase the raid size i used:
"mdadm /dev/md0 --grow--size=max"
now the array is reported as 1.5Tb.
"mdadm --misc -D /dev/md0"
however the underlying partition of fs (not sure which) is still at 750gb.
I created a test machine on a virtual version of fedora and tested out my theory before trying.
So my next step is to e2fsck -f /dev/md0 o check the array. once completed i need to umount the array then do a
"resize2fs /dev/md0"
however i cant unmount the array. I get "invalid argument". all the commands on the theucs are sym links to busy box. It has been reported as a bug, as noted above and on their site.
So i have no where to go from here presently until i can get the array unmounted. Just thought i would share what i have learned to date.
|
|
