Raid Lag

Discussion in 'Developer Roundtable' started by ARCHIVED-Silius, Aug 18, 2012.

  1. would love to raid test Hobarts new build on test server, lets find some peeps to put it together , am game for most time slots .
  2. could always set up a raid to go after highborn silius in leth... i am sure lots of people would love to log on spank that mob around with the gms in raid and watching.
  3. lest we forget. We had to abandon a core attempt today after lag made it all but impossible to do. These were players, some of whom were on their first APW raid. From the very start there was lag as other mobs were engaged, but the core fight came to a total halt for up to 30 seconds several times. The hopeful raiders tried more than once but it was just too much to overcome. I didn't see anything in Hobart's notes that were relative to raid lag, so its time to push this thread back to the top again.
    regards,
    leavwiz
  4. well from what ive seen the overall fps and smoothness ( hitching) of the game has improved quite significantly. i think ive gained at least 5-10 fps during raids which im really happy with.
    The long game freezes are still here however, and it seems they have still not found the root cause of them.
  5. Even on 12 man targets like the Sisters trials, we are noticing long latency pauses of up to 30 seconds.
    Its just not on.
  6. Alimora@Telon wrote:
    The improvements and enhancements to the Code that Hobart has been making, have been having some noticable affect. but.. what I still honestly think it comes down to, is whatever was done in april to the server/network/whatever that he has been trying to "smooth" out to cope with code. that is the underlying cause.. and sadly it refuses to be acknowledged or reverted. Instead Hobart has been tasked with the job to make the game run as smooth as possible without undoing that change.
    *edit* - Also.. Silius on vacation or something? haven't seen a post from him addressing any of these issues in a while now?
  7. leavwiz wrote:
    Most notable thing about our failed Core raid attempts was, that the huge lag always occurred at the same time, after roughly 30 seconds into the combat. So it is probably tied to some special timed ability of one of the present mobs, or one of the timed add spawns.
    Also, this lag was definitely server-software side. The client (at least mine) continued to draw more or less smoothly (as opposed to other situations that cause high client-side lag where the picture freezes), and the ping times to the chunk server that I was running in a separate window did not show significant changes. So it seems to be something happening that causes the server daemon to stop processing until it runs into a timeout and probably cancels whatever the mob/encounter was scripted to do..
  8. So....Should we just close all these topics out and forget this section of the forums every existed? Or is that going to happen anyway with the new forums. Because this doesn't seem constructive any more.
  9. I don't want you to think that this issue has been ignored, and our attention to it won't go away when the new forums come around. I know it's frustrating--we're also frustrated that we haven't been able to resolve the problem. And I know you're tired of talking about it, but please understand that I'm still trying to get my head around all the variables here.
    Just to be 100% clear, can I assume that the last two updates had no effect on the "fight comes to a halt for 30 seconds" issue?
  10. the fun pauses are less frequent yet still occurring
  11. over the last few raid nights we have experienced the long lag spikes several times during each fight. Although the occurances have not had the devastating effects of previous weeks (it appears that the npc stops doing whatever he is doing as well, so it is just a worrisome delay and timers get off). Previously we had more issues with raiders being locked up and some elements of the raid encounter still proceeding.
    We had lag spikes on Sparkles last night (even while clearing the trash mobs)
    We had lag spikes on War Golem but not a factor in the kill
    We had lag spikes on Sunshine (thanks for the spawn change) but didnt affect the outcome.\
    In the previous week lag was not a factor in the Shendu fight .
    There was lag on Masuke and Belzane, but those fights can be powered through the lag.
    There was lag on Teraxis resulting in a few deaths from bad timing on the enchantment phase and movement requirements.
    There was lag at Zodafin both before engaging and during the fight, but no complete lockups.
  12. Gotcha, thanks for being so specific.
  13. the fights still freeze sometimes. both the frequency and the duration have been been reduced since 3 months ago though. Also the general fps during raids has gotten a noticeable boost with your patches.
    Im not sure how it is related but in general the spikes seem to occur on the same day, then the next day can be completely smooth.
    What makes it bearable is that its the whole server freezing, including mob abilities and attacks, so if you spam your heal buttons you should heal as soon as it unfreezes, and thus covering the mobs dmg when it unfreezes. Downside is that timers like QJ and dmg boosters do keep ticking, so a lot of missed burns here and there.
  14. Alimora@Telon wrote:
    I too have noticed this. The framerate seems greatly improved from the past, it is much appreciated. If a solution to the 9999 ping spikes could be developed, raids would be smoother than they ever were in the past.
  15. Ping on Shendu and Karax seems significantly improved after the 1 Nov patch.
  16. Any further evidence after a patch, even anecdotal, is always useful. We don't call out every single optimization code change, because some of them would be hard to explain and it's tough to anticipate how much of an impact in a raid situation that each will have. But every patch since I've been on the team has included optimizations and memory leak fixes, including the 11/1 update.
  17. well my client hasnt crashed in 3 weeks, im quite happy with that :p
  18. Alimora@Telon wrote:
    The one and only time i logged on in the last 3 weeks and i crashed after using a riftway. Ironically it occured just after calling Alimora a noob.
  19. Yeah i hooked a d3d crash command in my reply :p

Share This Page