Here is how to get lvm and hal to leave your dmraid drives alone
James Olson
big_spender12 at lycos.com
Fri Dec 23 05:34:47 UTC 2005
The problem with running dmraid on Fedora Core 4 is that lvm and hal
look at the individual drives in the raid, do a partition scan, and
seek past the end of the drive with the big partition table on it.
Here is how to get them to stop doing that.
The following patch to /etc/lvm/lvm.conf causes lvm to ignore
drives /dev/hde, /dev/hdf, and /dev/hdg which are my dmraid array
drives.
--- lvm.conf~ 2005-04-27 12:32:08.000000000 -0700
+++ lvm.conf 2005-10-30 20:42:06.000000000 -0800
@@ -37,8 +37,8 @@
# that the cache file gets regenerated (see below).
# By default we accept every block device:
- filter = [ "a/.*/" ]
-
+ # filter = [ "a/.*/" ]
+filter = [ "r|/dev/hde*|", "r|/dev/hdf*|", "r|/dev/hdg*|", "a/.*/" ]
# Exclude the cdrom drive
# filter = [ "r|/dev/cdrom|" ]
The following filter file is named storage-raid-ignore.fdi and
can be added to /etc/hal/fdi/policy to
get the hardware abstraction layer daemon (hald) to ignore
those same drives.
<?xml version="1.0" encoding="ISO-8859-1"?> <!-- -*- SGML -*- -->
<deviceinfo version="0.2">
<device>
<match key="storage.bus" string="ide">
<match key="block.device" string="/dev/hde">
<merge key="storage.media_check_enabled" type="bool">false</merge>
<merge key="storage.no_partitions_hint" type="bool">true</merge>
</match>
<match key="block.device" string="/dev/hdf">
<merge key="storage.media_check_enabled" type="bool">false</merge>
<merge key="storage.no_partitions_hint" type="bool">true</merge>
</match>
<match key="block.device" string="/dev/hdg">
<merge key="storage.media_check_enabled" type="bool">false</merge>
<merge key="storage.no_partitions_hint" type="bool">true</merge>
</match>
</match>
</device>
</deviceinfo>
--
_______________________________________________
Search for businesses by name, location, or phone number. -Lycos Yellow Pages
http://r.lycos.com/r/yp_emailfooter/http://yellowpages.lycos.com/default.asp?SRC=lycos10
More information about the Ataraid-list
mailing list