Serviceguard NFS Toolkit A.11.11.06, A.11.23.05 and A.11.31.05 Administrator's Guide HP-UX 11i v1, v2, and v3
Table Of Contents
- Serviceguard NFS Toolkit A.11.11.06, A.11.23.05 and A.11.31.05 Administrator's Guide
- Table of Contents
- 1 Overview of Serviceguard NFS
- Limitations of Serviceguard NFS
- Overview of Serviceguard NFS Toolkit A.11.31.05 with Serviceguard A.11.18 (or later) and Veritas Cluster File System Support
- Overview of the Serviceguard NFS Modular Package
- Overview of the NFS File Lock Migration Feature
- Overview of NFSv4 File Lock Migration Feature
- Overview of Serviceguard NFS with Serviceguard A.11.17 Support
- Integrating Support for Cluster File Systems into Serviceguard NFS Toolkit
- Overview of Cluster File Systems in Serviceguard NFS Toolkit
- Limitations and Issues with the current CFS implementation
- Supported Configurations
- How the Control and Monitor Scripts Work
- 2 Installing and Configuring Serviceguard NFS Legacy Package
- Installing Serviceguard NFS Legacy Package
- Before Creating a Serviceguard NFS Legacy Package
- Configuring a Serviceguard NFS Legacy Package
- Copying the Template Files
- Editing the Control Script (nfs.cntl)
- Editing the NFS Control Script (hanfs.sh)
- Editing the File Lock Migration Script (nfs.flm)
- Editing the NFS Monitor Script (nfs.mon)
- Editing the Package Configuration File (nfs.conf)
- Configuring Server-to-Server Cross-Mounts (Optional)
- Creating the Cluster Configuration File and Bringing Up the Cluster
- Configuring Serviceguard NFS Legacy Package over CFS Packages
- 3 Installing and Configuring Serviceguard NFS Modular Package
- Installing Serviceguard NFS Modular Package
- Before Creating a Serviceguard NFS Modular Package
- Configuring a Serviceguard NFS Modular Package
- Configuring Serviceguard NFS Modular Package over CFS Packages
- 4 Migration of Serviceguard NFS Legacy Package to Serviceguard NFS Modular Package
- 5 Sample Configurations for Legacy Package
- Example One - Three-Server Mutual Takeover
- Example Two - One Adoptive Node for Two Packages with File Lock Migration
- Cluster Configuration File for Adoptive Node for Two Packages with File Lock Migration
- Package Configuration File for pkg01
- NFS Control Scripts for pkg01
- NFS File Lock Migration and Monitor Scripts for pkg01
- Package Configuration File for pkg02
- NFS Control Scripts for pkg02
- NFS File Lock Migration and Monitor Scripts for pkg02
- Example Three - Three-Server Cascading Failover
- Example Four - Two Servers with NFS Cross-Mounts
- 6 Sample Configurations for Modular Package
- Index
NOTE: If you enable the File Lock Migration feature, an NFS client (or a group of clients)
may encounter a situation where requesting a file lock on the HA/NFS server and not
receiving a crash recovery notification message when the HA/NFS package migrates to an
adoptive node. This occurs only when the NFS client sends its initial lock request to the
HA/NFS server and then the HA/NFS package moves to an adoptive node before the FLM
script copies the /var/statmon/sm entry for this client to the package holding directory.
The probability of hitting this corner-case problem is not very high, because the SM file copy
interval is very short (by default, five seconds). The chances of an NFS client (or group of
NFS clients) sending its initial lock request (it must be the initial request, since this request
generates the /var/statmon/sm file) to the HA/NFS server and having the package migrate
within this same five seconds window are extremely unlikely.
If you encounter this situation often, reduce the copy time interval by setting the
PROPAGATE_INTERVAL parameter to a lower value.
17. Modify the STATMON_WAITTIME parameter, that represents the amount of time in seconds
to wait before starting the cycle that copies files from the /var/statmon/sm directory into
the holding directory. This allows time for rpc.statd to complete its file lock migration
process since rpc.statd is started just before this script. The default value of this parameter
is 90 seconds. An example for this parameter is as follows:
For nfs.conf generated on Serviceguard A.11.18
STATMON_WAITTIME 90
For nfs.conf generated on Serviceguard A.11.19
nfs/hanfs_flm/STATMON_WAITTIME 90
Configuring Server-to-Server Cross-Mounts (Optional)
Two NFS server nodes may NFS-mount each other's file systems and still act as adoptive nodes
for each other's NFS server packages. Figure 3-1 illustrates this configuration.
56 Installing and Configuring Serviceguard NFS Modular Package