PCNS with two UPS units and a mix of servers - shutdown setup?

[Anonymous user]

Hi all,

We've got a rack (that is still being populated with servers) but for now I've got 7 servers, 2 of which have single supplies, the others have dual. Three servers are ESX, the others are either Linux or Windows and each UPS has an NMC.

I have installed and configured PCNS on all the systems and at this time I'm not exactly sure what will happen in the event of a power outage. Obviously for systems with the single supply, I've done the single UPS install vs dual supply servers have the multiple UPS config. What I'm trying to understand is how does the UPS finally decide it's okay to let its output be turned off? I'm not sure I can predict the overall shutdown time needed due to the multiple ESX servers, how long each guest ends up taking to shutdown, etc...

What I have at the moment is all clients are set to shutdown the UPS after the server finishes. Well, if server 1 takes 10 minutes to finally get all its guests shut down but server 2 only took 50 seconds since it's a single server...what happens?

I'm just a bit confused by the interaction of the servers and the UPS. Short of pulling both plugs and trying to ride out a shutdown...is there a way to predict how it will all pan out?

Thanks.

[Anonymous user]

KMP,

I don't see a setting with that exact label, I'm looking at the UPS/Configuration/Shutdown screen and there is a Low Battery Duration and Shutdown Delay.

I've looked elsewhere, but don't think anything else comes close.

[TheNotoriousKMP_apc]

The field is Low Battery Duration. They change the fields in firmware versions, but it should be set as a "warning" so that you know when the UPS has X amount of runtime remaining, it will send the shutdown command.

[Anonymous user · Accepted solution]

Hi all,

We've got a rack (that is still being populated with servers) but for now I've got 7 servers, 2 of which have single supplies, the others have dual. Three servers are ESX, the others are either Linux or Windows and each UPS has an NMC.

I have installed and configured PCNS on all the systems and at this time I'm not exactly sure what will happen in the event of a power outage. Obviously for systems with the single supply, I've done the single UPS install vs dual supply servers have the multiple UPS config. What I'm trying to understand is how does the UPS finally decide it's okay to let its output be turned off? I'm not sure I can predict the overall shutdown time needed due to the multiple ESX servers, how long each guest ends up taking to shutdown, etc...

What I have at the moment is all clients are set to shutdown the UPS after the server finishes. Well, if server 1 takes 10 minutes to finally get all its guests shut down but server 2 only took 50 seconds since it's a single server...what happens?

I'm just a bit confused by the interaction of the servers and the UPS. Short of pulling both plugs and trying to ride out a shutdown...is there a way to predict how it will all pan out?

Thanks.