erroronline14
Regular Member
- Oct 16, 2009
- 321
- 926
just what I thought seperating the posting from the wait reply - cut out the morning crash ... however still quite a few errors posting via web.
will test abit more and figure out whats tripping it up.
I have 25 instances running posting on web
I had 25 wait and reply running on the other PC but cpu usage was too high so cut back to 15 .... seems 5-9% each instance.
Wait and reply consumes a lot more resources than the tweet poster. It is the heaviest module in TA. So you can use more instances if you are using reply generator->tweet poster method.
And also, there are lot of things happening in twitter search now w/c wait and reply depends on. There are small changes happening every now and then so there are some instability but we are catching up and continuously updating it.