On 15-01-02 08:47 PM, Charles Lepple wrote:
So if you manually start the dummy-ups driver after the other two have connected to upsd, are there any other errors?
Sorry, I hadn't had a chance to test that yet. That appears to work fine (and therefore validate the race-condition hypothesis). The dummy driver starts normally and the client reports information about the "real" UPS which is exactly what I wanted. I will need to schedule a power-cycle test to make sure, but I don't see why things shouldn't work as desired now.
Depending on how you feel about compiling your own binaries or packages, it is possible to mix-and-match certain components from different NUT versions. In particular, the latest upsdrvctl can easily launch drivers from the original Ubuntu package (the "maxretry" functionality is entirely contained within upsdrvctl), and you could divert the package's upsdrvctl binary so that the changes are not overwritten. Let us know if you would like to try that.
I'm only a transient admin for this hardware, so I need to make decisions based-on what's easiest to document/maintain for others. I will need to think a little about whether it's simpler to have a manual call to "upsdrvctl start ups" somewhere or install from source...
Thanks again for all your help Charles, Jon _______________________________________________ Nut-upsuser mailing list [email protected] http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/nut-upsuser
