eq.gimasoft.com

Last night I used GINA on Valley Raid. When the event was over, after the Herald died, I continued to get a couple of Audio triggers about shield going up and down. The shield coming up for just a short time near the end of the fight often happens when the Herald is almost dead. If I had to guess, I got the shield messages approx 20 seconds late.

I can only guess that there was some lag due to the sheer number of log lines to process. My guild can be a bit swarm pet happy.

I've never experienced this when using GTT. I currently have 175 triggers in GTT, and I imported all of those into GINA, all of them are active.

In fairness, my log file was approaching 200mb at this point. But I have used GTT for the remainder of the evening, without issue, with the same log file. I tend to archive my log file at the end of the month, rather than at a set file size. I have had bigger log files before, with GTT, without issue.

Our Herald fight generated 30695 log lines in 154 seconds.

Thanks for the details; I'll look into this. Would you be adverse to emailing me your 175 trigger export file from GTT? I can generate the log volume you listed below, but having the actual triggers would be useful for testing. I didn't have any delay issues when I was parsing raids, but I only had about 30-40 triggers active, so I want to see where I need to add some performance.

The size of the log file shouldn't be a problem -- GINA jumps to the end of the file when it starts monitoring it, and then just processes the new data that comes in from that point on.

Also, did you happen to have the trigger Match Log showing (Settings -> Display Match Log)? It has two columns -- Log Time and Match Time. The Log Time is the timestamp read from the log file line itself, and the Match Time is the actual time that GINA processed that line and found a match. Ideally, these two values should be the same, but if there is a delay in the processing you will see a discrepancy in these lines (ie, if it took 20 seconds from the time the line was written to the log file to the time GINA matched it, the Match Time would be 20 seconds later than the Log Time).

I'm still trying to reproduce this one... thought I had, but then realized my log spam tester just wasn't flushing data to my log.

My current test is taking a raid parse from the Cazic Thule fight and playing it in super fast forward (100 lines every 20ms, or roughly 300k lines / min) with 211 triggers active, and GINA is keeping up with no delay. This is on a quad-core system with log files stored on a RAID5 array, so I'm going to build a dual core with normal hard drive this week and do some testing on that.

I did want to make sure of one thing, though -- when you say there was a 20 second delay, I assume there was silence during that time, right (ie, it wasn't a case of 20 seconds of Text To Speech getting queued up while other Text To Speech was running, right)?

Yes, there was 20 seconds of silence, then I trigger went off for shield up, and shortly after shield down. Both of which output text to speech for me.

My system is an Intel core2 quad Q6600 @ 2.40 GHz, with a single 3Gb/s sata drive.

I will test again next week, and see if it happens again. I didn't take note of trigger match time, sorry.

I was able to recreate the issue on an SSD drive. The issue is corrected in the latest version, 1.0.11.7. Thanks for reporting this!