Trakanons Teeth

[Brocode]

Is the zone ok, any know bug? I keep getting ping slowy going through 400 2500 4700 chars screen.

[OMGWTF420]

that zone is always laggy as hell

[Brocode]

looks like you have to zone in and out until you dont get hit by lag or something

[Brocode]

Considering its a known bug, how come i cant find anything about it, just feels like is being overcrowded/populated or nuked by something creating a mega lag?

[Ele]

http://www.project1999.org/forums/showthread.php?t=101722
http://www.project1999.org/forums/showthread.php?t=102035
http://www.project1999.org/forums/showthread.php?t=32600
http://www.project1999.org/forums/showthread.php?t=91754
http://www.project1999.org/forums/showthread.php?t=89643
http://www.project1999.org/forums/showthread.php?t=90715
http://www.project1999.org/forums/showthread.php?t=84947

[Brocode]

gotcha, still no lucky solving the problem, thanks tho

[nosto]

My guess is that the server for that zone is on a hard disk that is starting to run into issues on the SAN or is just starting to poop out and it creates "latency" as we see it on our side. We were in EJ last night and had massive packet loss for about 5 mins, then things simmered down.

[Thulack]

My guess is that the server for that zone is on a hard disk that is starting to run into issues on the SAN or is just starting to poop out and it creates "latency" as we see it on our side. We were in EJ last night and had massive packet loss for about 5 mins, then things simmered down.

I have no issues when i run through or hunt in TT. I think others can agree with me so doubt its a HD issue.

[MaximiusM]

My guess is that the server for that zone is on a hard disk that is starting to run into issues on the SAN or is just starting to poop out and it creates "latency" as we see it on our side. We were in EJ last night and had massive packet loss for about 5 mins, then things simmered down.

Really? a SAN? I bet the entire server runs on couple of Mac Mini with SSDs max max'd ram. Considering that a decent PC in 1999 had what, 16-64mb of ram and maybe a 20gb hard drive? Theres what, 72 zones right now? You could dedicate 1thread and 1GB ram per zone on 5 Mac Mini's and still have 4 cores and 8GB of ram left for load balancing and have spent less money than the SAN alone would cost... and thats paying the Apple premium.

Now, if the server was low on ram that could cause faults or high processor utilization could cause a thread crash which would translate into latency because the server is only processing data in fits...

[pharmakos]

i don't notice any connection lag, but TT is one of the only zones that my system lags on.... when i first zone into TT my framerate gets so bad that i can almost count the FPS.

[Byrjun]

This isn't framerate lag.

It only happens randomly, sometimes you zone in and you're fine, other times you're fine for maybe 10-20 seconds and then you start getting desynced. Things can still see and attack you so the server tracks your movement, but you can't see anything, you can't /con, can't send tells, etc. The only way to fix it is to zone (which will crash you) but Sebilis won't work as the server won't recognize you clicking the portal. You have to /q and come back and click the portal before you start getting desynced (once you get desynced it'll continue to occur until you zone somewhere, /camping or /qing won't fix it).

It's known and really needs to be fixed, especially since so many people are bound there to get to seb quickly (and you can't bind in seb).

[killa0885pve]

I made a thread about this in Bugs section, and pretty much got told its a client side issue, which i said was bullshit.... Thread ended there.

[Swish]

Also made a thread in the bug section.

1000 dissatisfied customers can't be wrong, amirite? Guaranteed if it was happening in EC it would have been fixed, but TT doesn't see too much traffic I guess :p

[Pend]

Really? a SAN? I bet the entire server runs on couple of Mac Mini with SSDs max max'd ram.
Or, as is more likely considering the difficulty in getting a solid i-net uplink in the above theory otherwise, they're running on a cloud server and impacted by unpredictable shared resource contention, which would cause unstable performance when pushed.

[MaximiusM]

I assumed against a cloud provider because the expense run rate would be pretty bad. that ish aint cheap

[Swish]

I'm going to go with "shit needs fixing in TT"...why is that the only zone to do it? It's got to be a bug/anomaly. It's single-handedly prevented me from joining a group in Sebilis a couple of times, and as a cleric they weren't impressed but all knew of the issue at least or had experienced it themselves.

I tried restarting, I tried deleting the TT files, I tried turning my monitor upside down, moving my mouse and keyboard outside of the room... it's not something on our side, its got to be server side.

[Morbo the Annihilator]

In no other zone do I experience the problems I have with TT.
There is no graphic lag, FPS is fine.
Ping skyrockets though, and I go LD at least 3 times before reaching seb.

When I finally limp into seb tunnel, clicking the orb is followed by ~30 second delay before I actually zone.
Once I'm in seb, life resumes as normal.

Twice in the past week, I've come back dead after a TT LD.

Definitely not a client issue.

[Sebekkha]

I had a problem with TT handing in my monks final turn in, I rebooted my client and all was fixed, albeit with very sweaty palms !!

[nosto]

Sorry, said san but it could just be the one servers drives, or a NAS it could be a lot of things. I had spoken to Rogean about the environment he is running the server on a while back, I know its virtual but the fact that its a specific zone and from my limited understanding of running an EQ server - either he has multiple VMs and the VM hosting the trak instance is also hosting another extremely volatile zone like OT or EC or it is just deprioritized period due to complete lack of people.

[Pend]

I assumed against a cloud provider because the expense run rate would be pretty bad. that ish aint cheap
I had a pretty decent cloud server for $100/mo. But yeah, the i-net traffic is usually metered. But keep in mind we're playing a game that was released when we all had 56k modems, so I can't imagine per user traffic to be that egregious.

Not sure if it's related, but I noticed insane (5-10s) lag every time I tried to use Sense the Dead in LOIO on the weekend. I tried it several times with the same result. I haven't used it since.

[MaximiusM]

Find an undead mob first and then cast sense, I bet there it is different. My assumption (totally guessing here) is that there is a client side buffer that precaches local (models, terrain, etc) data for a set distance beyond the clip plane. That buffer queries the server to preload mob info, path data, etc and serves the client as soon as the object is in plane.

What happens when something is father away than the cache radius? For speed the client should query the server directly but I doubt they build that logic for sideline cases. My guess is the client still has to query the cache, cache has to fetch, and then return to client. Enter your lag...

Same stuff happens if you bind at tunnel, run to t2, then gate back. Client has to repop all those models and their state, but that's way because they are in cache. Try the same experiment from Orc 2 and you'd probably get a longer delay.

This is my theory on why bards always look so laggy to me.