AST_cleanup_agent_log

[khuff]

Hey all. So I discovered the magic of the AST_cleanup_agent_log script today. We had a report [Agent time detail] that was ran yesterday evening after the agents got off and then again this morning and some of the agent talk times were WAY different. Like hours of talk time gone. Some agents went from 80% talk time to 20% talk time. In fact only one agent had a change of less than 5% in talk time, everyone else was between 10-60%. This has thrown everything into disarray as they were using the report in the evening to track performance, and come to find out that it is very inaccurate.

So I have two questions;

• Why is it off by so much? I get that it's doing some updates due to time differences, and fixing javascript time to server time, but I don't see where such a drastic change could be coming from. The log from yesterday had 40k updates.
• Can I run the 24 fix more than once a day? I don't ever see any output if I run the hourly one by hand. We have a master slave db setup, and they are pretty beefy so I'm not too concerned about load, just wondering if it'll throw stuff off more by running it more than once every 24 hours.

As usual any assistance is greatly appreciated.

[mflorell]

The main issue causing the initial inaccuracies is in Javascript. There isn't much we can do with that, so we had to write a process to correct those calculations. Yes, you can run that script more than once a day if you want.