Several of the SAP scripts have been written with a timeout feature to allow hung scripts to automatically kill themselves. This feature is required due to potential problems with unavailable NFS shares. This is explained in greater detail in the NFS Mounts and su topic. Each script equipped with this feature has a default timeout value in seconds that can be overridden if necessary. To reduce timeout wait times add required NFS file systems to ‘SAP_NFS_CHECK_DIRS’ as a comma-separated list. This enables ping to verify the file system is exported and if not it returns immediately.
SAP_CREATE_NAS can be enabled or disabled. The default for SAP_CREATE_NAS is 1 (enabled). This is used for automatically including a NAS resource for NAS mounted file systems. To disable, set this parameter to 0.
The default for the tunable SAP_CONFIG_REFRESH is LKCHECKINTERVAL/2. The user can change the number of seconds between calls to refresh the properties panel for an SAP resource in the LifeKeeper GUI.
The table below shows the script names, variable names, and default values. To override a default value, simply add a line to the /etc/default/LifeKeeper file with the desired value for that script. For example, to allow the remove script to run for a full minute before being killed, add the following line to /etc/default/LifeKeeper:
SAP_REMOVE_TIMEOUT=60
restore, recover | SAP_START_WAIT | 228 seconds |
remove | SAP_STOP_WAIT | 228 seconds |
remove | SAP_REMOVE_TIMEOUT | 804 seconds |
restore | SAP_RESTORE_TIMEOUT | 516 seconds |
recover | SAP_RECOVER_TIMEOUT | 1320 seconds |
quickCheck | SAP_QUICKCHECK_TIMEOUT | 60 seconds |
create NAS | SAP_CREATE_NAS | 1 (to disable, set to 0) |
GUI Properties Panel refresh | SAP_CONFIG_REFRESH | 1/2 the value of LKCHECKINTERVAL |
NFS shares to check | SAP_NFS_CHECK_DIRS | empty |
Post your comment on this topic.