NFS volume cools down when not accessed

[MrFilipot]

Hi,

I'm facing a curious issue - I mount an NFS volume (ONTAP 9.4P5) to RHEL 7.6 or CENTOS 7 and if I leave it as it is and later (after 1h) try df -h it takes 1-2 minutes (and strace shows it hangs on the NFS volume). After that all commands targeting the NFS just work as they suppose to - $time "real   0m0.009s". When I leave it to "cool down" for an hour the problem reoccurs again.

There is no FW, storage is accessed on L2. 

Workaround with df -h (or ll or whatever) running in the background periodically to keep the NFS "hot" is not an option.

Flags:rw,relatime,vers=3,rsize=65536,wsize=65536,namlen=255,hard,proto=tcp,timeo=600,retrans=2,sec=sys,mountaddr=10.203.205.244,mountvers=3,mountport=635,mountproto=tcp,local_lock=none,
addr=10.203.205.244

Thoughs?

[Vijay_ramamurthy]

Hi,

The issue reported seems to be curious indeed. 

I would suggest opening a support ticket so that support can review the packet traces and strace output to confirm if there are any issues with the NFS server.

Also curious to know whether the export rules are created using IP's or they are created using hostname/FQDN  or using netgroups.  

[MrFilipot]

Hi,

I will create the support ticket and paste the solution here when available.

---

@Vijay_ramamurthy wrote:

Also curious to know whether the export rules are created using IP's or they are created using hostname/FQDN  or using netgroups.  

---

Export rules are created using IPs.

[Vijay_ramamurthy]

Thanks.

If possible please share me the case number once created .  I am curious to take a look at the traces and ASUP logs myself  and understand the behaviour.