[EN] Today's patch also included another batch of small performance tweaks. Please let us know whether performance changes for you...no matter which direction.
[DE] Der heutige Patch enthielt auch wieder ein paar kleinere Performance-Optimierungen. Bitte lasst uns wissen ob sich die Performance für euch verändert hat...egal in welche Richtung.
Das ist soweit ok. 2 Minuten sind nicht wirklich "Rückstand". Viel interessanter ist, ob das Spiel für Spieler flüssig läuft. Denn das sollte es nach Möglichkeit immer, egal ob die Simulation im Rückstand ist oder nicht.
Stapleton hat einen Backlog von ca. 30 Minuten und die Reaktionszeit ist z.T. extrem lange
Das ist bekannt. Aspern und Stapleton teilen sich eine physische Maschine und haben sich gegenseitig etwas runtergezogen durch ihr Backlog. Sollte besser werden sobald beide wieder auf dem aktuellen Stand sind.
Stapleton has been sluggish for me basically since the spring cleaning in April, with very few exceptions. :angry: Today was particularly annoying with long lags. This applies to all write operations, not so much to read operations.
Sorry to say, but Aspern is extremely slow again (only 1 minute backlog), I find it very frustrating to do flight sheduling or route management... Is the hardware still the problem on this server?
i was scheduling some flights at gatow and again i would like to share some observations:
some operations went fast, some took some more seconds.
Everytime it took more than a second i watched the status page and i was surprised to see several times a green status pages with the next operation some seconds ahead.
While the site with the scheduling operation was still loading, i could load a fleet overview and confirm (by maintenance ratio shown) : the operation was already completed. Nevertheless it took more than 10 seconds to load the site, confirming the operation, status page all green.
for debug information:
ping
34 packets transmitted, 34 received, 0% packet loss, time 33040ms
rtt min/avg/max/mdev = 30.761/32.554/34.367/0.994 ms
traceroute
1 192.168.1.1 (192.168.1.1) 8.263 ms 8.234 ms 8.309 ms
2 151.23.227.139 (151.23.227.139) 9.805 ms 11.957 ms 14.231 ms
3 10.0.41.1 (10.0.41.1) 17.362 ms 19.774 ms 22.068 ms
4 151.6.234.73 (151.6.234.73) 24.195 ms 26.879 ms 29.563 ms
5 151.6.6.206 (151.6.6.206) 44.592 ms 151.6.6.202 (151.6.6.202) 34.186 ms 151.6.6.206 (151.6.6.206) 44.704 ms
6 151.6.1.190 (151.6.1.190) 50.892 ms 151.6.1.130 (151.6.1.130) 21.981 ms 24.098 ms
7 151.6.3.66 (151.6.3.66) 20.077 ms 151.6.1.142 (151.6.1.142) 24.635 ms 151.6.1.202 (151.6.1.202) 26.144 ms
---- until here my provider
8 tge-5-1-0-353a.cr2.fra.routeserver.net (80.81.192.21) 37.737 ms 38.795 ms 42.524 ms
9 217.118.16.26 (217.118.16.26) 58.496 ms 48.728 ms 63.023 ms
10 217.118.16.166 (217.118.16.166) 64.575 ms 217.118.16.164 (217.118.16.164) 67.773 ms 217.118.16.166 (217.118.16.166) 59.368 ms
11 loft2536.serverloft.de (85.25.236.90) 63.831 ms 65.304 ms 34.138 ms
The behavior described by andreamilano does indeed sound a bit strange. I'll add some debug code later today and apply a small patch to Gatow to figure out why this happens.
I have applied a small patch to gatow to display a few measurements about the execution times of writing operations.
So when you, for example, apply new schedule settings, something like this will show up in the footer:
stats (i/e/o/t): 1/18/0/19
If your execution times feel longer than they should, please post these figures here. Since I can't reproduce most of the performance issues in vitro, this will help me to gather some insights.
Aspern was running very well this morning (around 8:30 AM till I don't remember, sorry). Is there a chance that maybe in the evening (i.e. around 6:37 PM) more players are active (or more traffic in general)? Because now it takes quite some time to check my route management and flight scheduling. I'm no expert in all this, just got me thinking, because in the morning planning was - for the first time in a long time - fun again. Sorry if this is not helping at all :) .