[Nut-upsdev] Getting 'Data stale' error with bcmxcp_usb for a PowerWare 5115 on OSX
Charles Lepple
clepple at gmail.com
Mon Mar 15 01:39:46 UTC 2010
On Mar 14, 2010, at 4:32 PM, Arjen de Korte wrote:
> Citeren Charlie Garrison <garrison at zeta.org.au>:
>
>> And I answered my own question (leaving it here in case it helps
>> someone else later). I re/plugged the UPS USB cable and the driver
>> died. When/if I get the 'data stale' error again I'll send anything
>> from the debug log that looks relevant.
>
> I think this is what needs to be solved. If the driver would just
> reconnect to the UPS whenever the connection is lost (either by
> fork'ing to the background or because the USB plug was replugged),
> you'd probably not have noticed anything in the first place. Any USB
> driver that not automatically reconnects is to be considered broken
> (temporary loss of connection is all too common with USB).
Arjen,
two things here:
1) I agree that the bcmxcp_usb driver should try to reconnect if it
sees an event which can be solved by reconnecting. Maybe we can apply
the same logic as on the other USB drivers, however...
2) It seems that there are still some problems with USB connections
after forking - I tried to get the tripplite_usb driver to reconnect
(and it has been modified to do this under Linux) but I saw
inconsistent results on OS X.
Who is in charge of bcmxcp_usb these days?
--
Charles Lepple
More information about the Nut-upsdev
mailing list