mherger wrote: > > So this morning there was this pretty big DSL blackout in Germany > caused > > by a storm. I noticed that in the log LMS announced it would retry > > connecting to MSB in 86400 seconds. That's 24h hours. But instead, > LMS > > retried it 30s later. There seems to be something wrong with the > numbers > > in that particular error: > > There must be a bug around the mysb.com sign-in I still haven't found. I > > see from mysb log files that sometimes an instance would hammer the > service with thousands of login requests in a very short time (millions > > within a day!). The LMS side back-off strategy in this case probably > would fail, as more requests might have been triggered before the first > > error comes back. > > But going to weeks in between re-tries seems wrong, too... > > -- > > Michael
My LMS really was offline for several hours, not blocked by MSB. I do wonder why the claimed retry interval would be increased by a constant every retry while the actual interval doesn't seem to follow some easy to spot logic ([s]: 30, 50, 46, 153, 98, 105, 300, 4076(!), 321). The log shows two instances of time where it's hh:mm:ss.*1*xxx. That's where the 4076s spike is. Could be coincidental, of course. Then again, you'd need milliseconds to achieve your millions of login requests a day. QLMS [email protected] x64 (digimaster) with perl 5.28 dedicated to me. :D / QNAP 469L QTS 4.3.4 ------------------------------------------------------------------------ dolodobendan's Profile: http://forums.slimdevices.com/member.php?userid=67663 View this thread: http://forums.slimdevices.com/showthread.php?t=111656 _______________________________________________ Squeezecenter mailing list [email protected] http://lists.slimdevices.com/mailman/listinfo/squeezecenter
