Coyote Point Systems E350GX, Installation And Administration Manual

Page: 306 / 344

Appendix F: Equalizer VLB
302 Equalizer Installation and Administration Guide
7. In the
functions
field, click
running
. In the
expression workbench
field at bottom, click on the drop-down
arrow next to
running
. Select
sv01
and click
accept
. The
expression workbench
should now look like this:
8. Click the next icon (
>
)at top to open the
Event Action
editor.
9. In the
functions
field, click
quiesce
. In the
expression workbench
field at bottom, click on the drop-down
arrow next to
quiesce
. Select
sv00
and click
accept
.
10. In the
operators
field, click
&&
.
11. In the
functions
field, click
quiesce
. In the
expression workbench
field at bottom, click on the drop-down
arrow next to
quiesce
. Select
sv01
and click
accept
.
12. In the
operators
field, click
&&
.
13. In the
functions
field, click
unquiesce
. In the
expression workbench
field at bottom, click on the drop-down
arrow next to
unquiesce
. Select
sv02
and click
accept
.
14. In the
operators
field, click
&&
.
15. In the
functions
field, click
log
. In the
expression workbench
field at bottom, click on the drop-down arrow
next to
log
and type “Last hot spare activated! Check servers!”. Click
accept
. The
expression workbench
should now look like this:
16. Click the next icon (
>
)at top, and then
commit
to create the event. The Configuration tabs for the event open in
the right frame.
Note that the above example is very basic and does not handle all possible event combinations. In particular:
• No action is taken when one of the hot spares becomes unavailable while the first server is passing traffic.
These will be indicated as down in the Cluster and Server tabs.
• No action is taken if the third server is not available. This could be handled by a separate event.
• No action is taken if all three servers are not available. This is best handled by a Responder combined with
a Match Rule.
Rebooting an Unresponsive Virtual Machine (VLB only)
In this example, we want to reboot a particular virtual machine running under VMware when the machine has been
unresponsive for 15 minutes.
To do this, we create two events to power up and power down the machine, and a third event that acts as a timer for
the power down and power up events.
Event 1:
If the server is not running, power the server down; otherwise, do nothing.
Event 2:
If the server is not powered on, power the server on; otherwise, do nothing.
Event 3:
If the server is running, block Event 1 and Event 2 for 15 minutes; otherwise, do nothing.