RHEL5.8: NFSv4 operations intermittently result in I/O errors, specifically lost WRITE's
Issue
- Informatica session logs indicate IO errors occur periodically
- IO errors often occur during peak times of 1am to 5am
- Periodically NFS transactions never make it to the NFS server and the client states it cannot connect to the server:
Oct 19 04:08:38 hostname _k_e_r_n_e_l_: nfs: server nfs-server not responding, still trying
Oct 19 04:08:38 hostname _k_e_r_n_e_l_: nfs: server nfs-server not responding, still trying
Oct 19 04:08:44 hostname _k_e_r_n_e_l_: nfs: server nfs-server not responding, still trying
Environment
- NFS Client
- Red Hat Enterprise Linux 5.8 kernel (5.7 system)
- kernel 2.6.18-308.8.2.el5, contains fix for NFSv4 client hang while database write load simulator running
- NFS Server
- NetApp: Ontap 8.02P3 (7-mode)
- Mostly NFSv4 but NFSv3 also used to mount some exports from the same NFS server
- Specifically the top level mount is an NFSv3 mount. Underneath the top level mount are many additional NFS mounts that are both V3 and V4.
- Application which does repeated sequences of the following NFSv4 operations
- OPEN
- WRITE
- CLOSE
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.
