SLES 15.0 Pacemaker NFS and autofs dont work together


Running on Azure
Setup nfscluster for exposing nfs mounts.
This works

Next, setup SAPcluster for running central SAP services.
Using the azure guides.
First mount the NFS mount which do not come from the local pacemaker cluster.

Map the nfs
vi /etc/auto.master
Add the following line to the file, save and exit
/- /etc/

vi /etc/
/sapmnt/SID -nfsvers=4,sync AZURELBNFS:/SID/sapmnt
/usr/sap/trans -nfsvers=4,sync AZURELBNFS:/saptrans
/usr/sap/SID/SYS -nfsvers=4,sync AZURELBNFS:/SID/SYS

Starting autofs and the mounts are available.

Now setup pacemaker, which mounts ASCS from the same NFS cluster on the SAPcluster:
sudo crm configure primitive fs_SID_ASCS Filesystem device='AZURELBNFS:/srv/nfs/SID/ASCS01' directory='/usr/sap/SID/ASCS01' fstype='nfs4' \
op start timeout=60s interval=0 \
op stop timeout=60s interval=0 \
op monitor interval=20s timeout=40s

sudo crm configure primitive vip_SID_ASCS IPaddr2 \
params ip= cidr_netmask=27 \
op monitor interval=10 timeout=20

sudo crm configure primitive nc_SID_ASCS anything \
params binfile="/usr/bin/nc" cmdline_options="-l -k 62001" \
op monitor timeout=20s interval=10 depth=0

sudo crm configure group g-SID_ASCS fs_SID_ASCS nc_SID_ASCS vip_SID_ASCS \
meta resource-stickiness=3000

If I now start the resource, the autofs filesystems timeout or become really slow.
The NFS mount from pacemaker can be used as it should.
If I stop the nfs resource in pacemaker, my autofs mounts work nicely again.

What is happening here?

Thanks for you help


  • strahil-nikolov-dxcstrahil-nikolov-dxc Established Member
    What happens when you stop the cluster and manually mount the NFS (as if it was done by the cluster) ? Does autofs get slow ?
Sign In or Register to comment.