Without a doubt, things are so much better than when we started on this merry journey nearly 2 years ago. Frame rates are up, network lag is down and things are just peachy, most of the time. Yet, we still have those occasional situations where the game becomes a bit of a nightmare. You’re stuck in endless loops of effects, people aren’t where you’d expect them to be, your enemy rides past you, assuming that you’re not really there, the CTDs start to appear and spell effects appear in clusters.
On the CTD side of things, this game is uber stable for me, apart from when the numbers get very high and then you end up staring at the desktop. It’s that age-old problem, that running any engine at max for to long and the seams start to split, which in this case maybe a small if then else, where the else is “Oh Shit, To Much Lag = Dump User To Desktop aka Clear The Buffers”.
I experienced such an evening last night 24/08/10, while on Karak-Norn in Eataine. It got so annoying, that I went and did what I hate most, a bit of crafting. Maybe because it’s such a rarity for me to experience issues like these, I find it that bit more frustrating. It could be that I’m just an intolerant old fool who values his free time more as he gets older, and doesn’t want it affected in anyway.
With the time afforded to me by the crafting, it allowed me to ponder a couple of things. Could the technical setup we have, both from a hardware and software standpoint, have handled Warhammer being really successful, as many of us had originally hoped for? We would have had more servers, this is true, but those end zones would have been mighty packed.
As we know from recent announcements by Mythic, design decisions are still being made in an effort to reduce the load on the server. In this case it was collision detection, which got shelved after a public outcry. Since Mythic hold their cards close to their chest re: hardware, we don’t know much about their kit. I’ve always assumed it was a blade server running each of our virtual servers (Karak Norn for instance), with each blade holding a region(s). What I’m not sure about, is how threaded is the server-side code when aligned with the hardware it’s running on. This does lead me to believe, that no new hardware will be forthcoming, though as usual I’m probably talking out of my ass.
When I started writing this, it got me thinking about why this push on performance? Is it just another step in fight against server side lag or is it because bottlenecks are starting to appear more frequently, because heaven forbid numbers starting to climb again? Maybe the new RvR packs and the appearance of the Skaven in some form as yet undisclosed could spike the servers. Remember about a month ago, we had a free 2 week trial on the Euro servers and the numbers climbed again, by what seemed quite a lot. Oh btw I’d love to know the number of renewed subscriptions based upon that PR exercise. Would the RvR packs produce a similar effect across all the servers, is that the worry?
Yes, it’s all speculation, but what else do I have to talk about. I know bugger all, naff all and sod all about the RvR packs.
Just for the record, hardware wise its a I7-920 @ 2.66, Asus P6TD Deluxe MB, 6GB Memory, ATI 5850, old school 300GB WD Raptor, sitting at the end of a 50MB/1.5MB internet connection (Cable, so it does the speeds it says on the tin), running at 1920 x 1280. I wanted to banish any thoughts that the client pc was underspecced.
EDIT: The more I think about it, the more I think it’s server processing overload. If it’s network traffic, it shouldn’t be a problem in another part of the same zone, since you should only be sent network traffic about things that effect you, so if you can’t see the ruddy great big battle, you’d not be sent any network traffic related to it, beyond the deathspam. However the only thing that would make the network seem sluggish while being in the same zone but not in sight, is if the processor(s) were struggling to cope with the battle in the zone and were therefore having to queue you request for mounting the horse, for example.
The crashes never really went away after release; I busy RVR I can bet on crashing within the hour, and not just there, I’ve crashed in the middle of LotD, far away from anyone. There seems to be no identifiable cause for this. Drivers all up to par, decent hardware, pc doesn’t even have to break into a sweat.
I’ve gotten used to it, but that isn’t really a good thing, is?
I rarely have CTDs, it’s really only when I am in very very large keep takes. Even very large battles don’t seem to have the same problem.
I think it’s more of an issue with the game client than it is with the servers themselves. Server-side, things seem to be hunky dory, tight and solid. The game client itself seems to have major problems when running into massive amounts of people. Which means the game itself needs to be optimized, and no amount of external hardware will really help with that.
I think this is most clear when you look at how often server crashes actually happen. I almost never see whole zones or servers go offline unexpectedly anymore. But I still crash on my end from time to time after fighting in huge battles for long periods of time.
The servers may not be crashing that often, which is a good thing. However that doesn’t mean they are able to handle the load any better than before, though of course optimisation is an ongoing process.
The issue of data arriving in bursts, like me getting the effect of 5 AOE’s effectively going off as one, though a sympton of network traffic, could be the fact that the processing of actual AOE and who it hits, are being bottlenecked on the server.
Of course we could be having issues, that the client maynot be handling the gaps in network traffic very well and effectively roles over and dies.
I have plenty of bandwidth to handle the traffic and any calculations/gfx my end. As Tesla said, who has a similiar high end pc, we have the grunt.
I still think it’s the server. Which is why Mythic considered dropping something quite fundamental to ORvR, which is collision detection. That wouldn’t have been on the table if the servers were fine with their loading. Which is also why a comments were made about the feel of the game on the PTS when it was out.
If I’m in a big fight it’s usually a CTD every half an hour. I have a feeling it’s just the game going “OK, this isn’t working let’s try again”. I hate it.
Btw Skar, I never knew you were on KN. What’s your nick there?
Shesk, you’ve never read my About then http://thatsskarwithak.wordpress.com/about/
At the moment, my main is Laminar, which is latin for flame.
Yeah, the moment I posted that I went to look for an about, didn’t bother rereplying (no edit on wordpress damnit!) because I just thought you’d ignore it.
Now Laminar, I’m sure I’ve grouped with you before, because that name is very familiar. Though you probably wouldn’t know it because I play so many chars. xD
I’v started to get CTDs since the last patch
not very often, but it’s new for me (well not entirely new, I played from release)
But I guess it could be bcouse of the numbers of players in t4 have increased…it’s just speculation.
Skar your PC is pretty good I think.
I run on
Intel core2 pro 2,66Mhz
4GB oldschool 400mhz RAM
5870 Radeon
7200spin 500GB HDD
24″ Monitor
You machine is good, with a better gfx card than mine.
Now you need to separate out the fact about the patch turning up and think more about the number of people who are playing.
We had the double rp bonus, the against all odds is working. So the CTDs could well be down to increased population. There may well be crux point, where the number of people in a zone can tip the blade into senile demensia.
@Blaq,
Fastrada I definitely recognise, and your RP as well.
Well don’t be stranger! I’m a love maching, and like to share.
The “I’m a love machine and like to share” worries me a little !!!
my comp sucks so im screwed either way.