From Bugzilla Helper: User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:0.9.2.1) Gecko/20010901 Description of problem: autofs is mounting the same file system twice. Check out the following df for an example. [ofer@g2pro2 2000-137]$ df Filesystem 1k-blocks Used Available Use% Mounted on /dev/sda1 1983964 1673132 208420 89% / none 256228 0 256228 0% /dev/shm /dev/sda3 6496640 4924064 1237220 80% /home /dev/sdb1 17063007 14790058 1384320 92% /local.dsk/g2prod2 physgi00:/users/usr0 97684664 82132896 15551768 85% /users/usr0 /local.dsk/g2prod2 17063007 14790058 1384320 92% /phyppro/g2prod2 phyppro2:/local.dsk/g2disk1 21702268 20355421 213865 99% /phyppro/g2disk1 physgi00:/users/usr0 97684664 82132896 15551768 85% /users/usr0 phyppro2:/local.dsk/g2disk1 21702268 20355421 213865 99% /phyppro/g2disk1 /local.dsk/g2prod2 17063007 14790058 1384320 92% /phyppro/g2prod2 Version-Release number of selected component (if applicable): How reproducible: Sometimes Steps to Reproduce: 1.load redhat 7.2 2.configure automount 3.what automout mount same mount point twice Additional info:
The relevant parts of the autofs configuration files on this system are as follows: auto.master: /phyppro /etc/auto.phyppro --timeout 60 /users /etc/auto.users --timeout 60 auto.phyppro: g2disk1 -nosuid,nodev,hard,intr,rw,rsize=8192,wsize=8192 phyppro2:/local.dsk/g2disk1 g2prod2 -nosuid,nodev,hard,intr,rw,rsize=8192,wsize=8192 g2pro2:/local.dsk/g2prod2 auto.users: usr0 -nosuid,nodev,intr,rw,rsize=8192,wsize=8192 physgi00:/users/usr0 and the mount command returns the following information (at a later stage now wherein the number of multiple mounts has increased): # mount /dev/sda1 on / type ext3 (rw) proc on /proc type proc (rw) none on /dev/shm type tmpfs (rw) none on /dev/pts type devpts (rw,gid=5,mode=620) /dev/sda3 on /home type ext3 (rw,nosuid,usrquota) /dev/sdb1 on /local.dsk/g2prod2 type ext3 (rw) automount(pid754) on /phyppro type autofs (rw,fd=5,pgrp=754,minproto=2,maxproto=3) automount(pid815) on /users type autofs (rw,fd=5,pgrp=815,minproto=2,maxproto=3) physgi00:/users/usr0 on /users/usr0 type nfs (rw,nosuid,nodev,intr,rsize=8192,wsize=8192,addr=130.199.22.67) /local.dsk/g2prod2 on /phyppro/g2prod2 type none (rw,bind) phyppro2:/local.dsk/g2disk1 on /phyppro/g2disk1 type nfs (rw,nosuid,nodev,hard,intr,rsize=8192,wsize=8192,addr=130.199.22.10) automount(pid2153) on /phyppro type autofs (rw,fd=5,pgrp=2153,minproto=2,maxproto=3) automount(pid2231) on /users type autofs (rw,fd=5,pgrp=2231,minproto=2,maxproto=3) physgi00:/users/usr0 on /users/usr0 type nfs (rw,nosuid,nodev,intr,rsize=8192,wsize=8192,addr=130.199.22.67) phyppro2:/local.dsk/g2disk1 on /phyppro/g2disk1 type nfs (rw,nosuid,nodev,hard,intr,rsize=8192,wsize=8192,addr=130.199.22.10) /local.dsk/g2prod2 on /phyppro/g2prod2 type none (rw,bind) automount(pid21046) on /phydsk type autofs (rw,fd=5,pgrp=21046,minproto=2,maxproto=3) automount(pid21068) on /phyppro type autofs (rw,fd=5,pgrp=21068,minproto=2,maxproto=3) automount(pid21097) on /physgi type autofs (rw,fd=5,pgrp=21097,minproto=2,maxproto=3) automount(pid21124) on /amd.aux type autofs (rw,fd=5,pgrp=21124,minproto=2,maxproto=3) automount(pid21148) on /users type autofs (rw,fd=5,pgrp=21148,minproto=2,maxproto=3) phyppro2:/local.dsk/g2disk1 on /phyppro/g2disk1 type nfs (rw,nosuid,nodev,hard,intr,rsize=8192,wsize=8192,addr=130.199.22.10) /local.dsk/g2prod2 on /phyppro/g2prod2 type none (rw,bind) physgi00:/users/usr0 on /users/usr0 type nfs (rw,nosuid,nodev,intr,rsize=8192,wsize=8192,addr=130.199.22.67)
Same problem, have the same directory mounted 24 times now after one day. $ mount|fgrep porky|wc -l 24 [harald@faro harald]$ mount|fgrep porky|sort -u porky.stuttgart.redhat.com:/mnt/raid on /mnt/raid type nfs (ro,rsize=8192,wsize=8192,nolock,soft,intr,bg,addr=172.16.2.140)
yeah, autofs has to better-handle existing mounts in its fs-space or we need a mechanism to selectively disable the "supermounting" feature - since the latter is unlikely, it will (not that it should) fall back to autofs to perform a check on the would-be mountpoint to make sure it's not already the target of an existing mount (a stat just checking to see if it exists, perhaps) before it attempts the mount. Until it does so, it'll keep happily remounting the same existing mount ad infinitum. *** This bug has been marked as a duplicate of 65003 ***
On my systems the number of multiple mount becomes scarily large. Has the problem ever been fixed?