|
Operating
SavWareHAError Conditions, Detection, & Recovery
Resolving SavWareHA Mode Discrepancies
After
both systems are booted to multi-user mode, administrator action
may be required if the SavWareHA modes get out of sequence between
the two systems. Refer to the following table:
| Primary
System |
Standby
System |
Actions
|
| Normal |
Normal |
None required.
The Primary and Standby System s are both in normal operation.
Mirroring continues. |
| Normal |
Fallback |
Communication
has been lost on all channels between the systems. The filesystem
data is out of sync. Both systems may have taken control of
their respective disk partitions. Restore communication and
follow the procedure(s) shown in ‘Unique or Difficult Situations’
section of this manual . |
| Fallback |
Fallback |
The Primary
System has or had failed. The filesystem and users are controlled
by the Standby System . The Primary System can be tested and
repaired while the users are running on the Standby System.
When the system administrator is confident that the Primary
System is operational. The "Recover to normal mode "
option of SavWareHA should be used. |
| Fallback
|
Normal
|
An unrecoverable
error has occurred on both systems. Neither system has control
of the users or disk partitions. The filesystem integrity may
be compromised. Test all communications links and follow the
procedure(s) shown in ‘Unique or Difficult Situations’
section of this manual . |
Switch
to Fallback Mode
In the case
of a Primary System failure, the Standby System 's monitor program
detects a failure by the lack of access acknowledgments across all
entries specified in the ‘/etc/sentinel.d/links’ file.
When this occurs, SavWareHA will automatically switch to fallback
mode -- if the keyword ‘manual ’ is not present in the
‘links’ file. When SavWareHA switches to fallback mode
either by the SavWareHA monitor program or manually from the SavWareHA
utility menu, the following events occur:
|
