Symptom
A modular / multi-slot model of ASR 1K router may experience various NFS (network file share) errors on either their CC (carrier card) modules (SIP and MIP modules) or on the FP (ESP engines). This will continue indefinitely and fill the syslog.
Some various known types of errors:
== SIP / MIP Errors ==
SIP0: kernel: NFS: server 10.0.1.0 error: fileid changed fsid 0:XX: expected fileid 0xX, got 0xXXXX
== ESP Engine Errors ==
F0: kernel: INFO: task mount.nfs:X blocked for more than X seconds.
Conditions
- Router is running an IOS-XE 16.9 .x or 16.10.x image.
- Seen with modular models of ASR 1K
- Seen with RP2 or RP3 engine
In this context of this issue, the SIP / MIP / ESP card is attempting to reach IP 10.0.1.0 which is an internal address to the RP engine which hosts an NFS (network file share) mount. Due to a timing issue, the NFS mount path can be lost under the following situations and will not attempt an automatic re-mount for recovery:
1) During a reload of both the RP engines.
2) During multiple RP failovers (eg: as can happen during an IOS upgrade)
3) During multiple FP failovers (eg: as can happen during an IOS upgrade)
Workaround
Reload the SIP, MIP, or ESP engine that is printing the NFS error logs. Eg:
ROUTER# hw-module slot reload
Further Problem Description
This bug will be fixed in IOS-XE 16.9.4. See "known fixed releases" section for more info.
The NFS link status needs to be verified by the following CLI:
sh platform software mount fp (cc)
