Please upgrade your browser for the best possible experience.

Chrome Firefox Internet Explorer
×

Instance Lag, Jedi Covenant

GOTGK's Avatar


GOTGK
12.06.2013 , 12:23 AM | #1
Hey folks,

Not sure how many other folks attempted to raid this evening (12/5), but the lag has been encounter breaking. Has anyone heard anything about a fix in the works for this new issue (since the star-farter pack launched and started eating up instances)?

I don't know how many cycles the dev team has given to their current flagship (heheheh) minigame, but it's affecting things rather badly for the rest of the instanced raid groups on my server. Entire raids are completely unresponsive for up to 15 or 20 seconds. Chat works ok usually, everyone's vent is fine, and here's the fun bit--- THE BOSS/EVENT/MOBS ALSO STOP DOING ANYTHING!

It's nice to come back from a massive spike like that and not be dead, but the fact that we aren't dead pretty much confirms that all of your cycles are going elsewhere lately. I can only imagine it's the SpaceFace.

Please re-evaluate your load balancing. I don't know if you're giving bandwidth priority to Space PeeVees, but it seems like you're prioritizing all of the instanced starfighter content on the processing front. I get that you want your new content to work well! (And for what it's worth, I'm not a fan of traditional PVP and the associated grinds, but I enjoy the approach you guys are taking so far w space PeeVee).

But at some point you have to stop throwing dirt on one part of the game to make another look shinier by comparison. Just cleaning up both seems easier, but then-- hey-- I'm not a bean counter, and I'm sure it's more expensive.

If instanced content is hosted on the characters' home server, that's fine. I don't know what resources you're working with. But I have a sneaking suspicion that your Space Face is set to have a massive priority on cycles right now, and you should know it's breaking the regular PVE raiders over here on Jedi Covenant.

We can only take the entire raid lagged out several times an hour so many times before we have to call the raid night. It's not like the raid is coming back to dead bodies, we're sitting there groaning in vent for 10-30 seconds (yes we timed that one (32 seconds)--average is about 14 seconds) and we come back to mobs who also have done nothing in our absence (but the timers keep ticking for enrage, etc-- that seems to be tied to a network/SQL table, not actively processed, per se) For mobs this is fine, they just pick up where we were, and we keep fighting. But on bosses, they always reset. Since the server processes it's data first, I'd imagine the boss on the server side finally gets his or her cycles and says, 'Yeah, I'm gonna pound that Juggernaut!' only to find that the client has yet to process it's own downstream data, where the Jugg is returning a phonecall from her boyfriend and still lagged out for a few more seconds.

End result? Well, you know, Boss gets cycles, suddenly finds nobody active on his or her table and resets. Clients gets their cycles, and suddenly our 24% boss has 100% HP again. YAY!

We had this happen 4 times on one boss before we went a miraculous 9 minutes without your hardware postponing our activity.

The problem is, I know this is a resource/priority issue, because I manage systems that host 750k-2million concurrent users. And I know it's tricky, so I'm not mad like my friends. I know all of the things you guys could possibly say in response to this issue because I've issued the same responses-- hell, I can't tell you how many canned responses I've had people record into switcher messaging systems. Sometimes the best you can say to the end user is, 'we're aware of the issue and workin on a resolution' even if corporate is only focused on one particular project.

SO! If you are hosting /all/ instanced activity on the user's home server, and I would guess you are, please stop giving the spaceface instances all of the damned cycles during heavy load. Either host them on another resource, find a better way to spread them across the VMs, or give them the same priority the rest of us get. There are about 3 different different ways I deal with this kind of thing when I have a new product that upstairs wants to shine, and I'm sure you guys know more about the specifics of making it happen than I ever will.

Just let it be known, it's breaking encounters. And we know why. If those that decided on this course of action discard our frustration, remind them:

Your company is paid to provide a persistent service. If Verizon's HD network turned my TV into frozen, pixelated garbage for 15-30 seconds about every 15 minutes, do you think I (our they) would accept that amount of loss? Even if it only occurred on 60% of the channels?

Pete looks great, but Paul and Mary are getting marginalized.

Sincerely,

Pete

Hatstandard's Avatar


Hatstandard
12.06.2013 , 03:03 AM | #2
Quote: Originally Posted by GOTGK View Post
Hey folks,

Not sure how many other folks attempted to raid this evening (12/5), but the lag has been encounter breaking. Has anyone heard anything about a fix in the works for this new issue (since the star-farter pack launched and started eating up instances)?

I don't know how many cycles the dev team has given to their current flagship (heheheh) minigame, but it's affecting things rather badly for the rest of the instanced raid groups on my server. Entire raids are completely unresponsive for up to 15 or 20 seconds. Chat works ok usually, everyone's vent is fine, and here's the fun bit--- THE BOSS/EVENT/MOBS ALSO STOP DOING ANYTHING!

It's nice to come back from a massive spike like that and not be dead, but the fact that we aren't dead pretty much confirms that all of your cycles are going elsewhere lately. I can only imagine it's the SpaceFace.

Please re-evaluate your load balancing. I don't know if you're giving bandwidth priority to Space PeeVees, but it seems like you're prioritizing all of the instanced starfighter content on the processing front. I get that you want your new content to work well! (And for what it's worth, I'm not a fan of traditional PVP and the associated grinds, but I enjoy the approach you guys are taking so far w space PeeVee).

But at some point you have to stop throwing dirt on one part of the game to make another look shinier by comparison. Just cleaning up both seems easier, but then-- hey-- I'm not a bean counter, and I'm sure it's more expensive.

If instanced content is hosted on the characters' home server, that's fine. I don't know what resources you're working with. But I have a sneaking suspicion that your Space Face is set to have a massive priority on cycles right now, and you should know it's breaking the regular PVE raiders over here on Jedi Covenant.

We can only take the entire raid lagged out several times an hour so many times before we have to call the raid night. It's not like the raid is coming back to dead bodies, we're sitting there groaning in vent for 10-30 seconds (yes we timed that one (32 seconds)--average is about 14 seconds) and we come back to mobs who also have done nothing in our absence (but the timers keep ticking for enrage, etc-- that seems to be tied to a network/SQL table, not actively processed, per se) For mobs this is fine, they just pick up where we were, and we keep fighting. But on bosses, they always reset. Since the server processes it's data first, I'd imagine the boss on the server side finally gets his or her cycles and says, 'Yeah, I'm gonna pound that Juggernaut!' only to find that the client has yet to process it's own downstream data, where the Jugg is returning a phonecall from her boyfriend and still lagged out for a few more seconds.

End result? Well, you know, Boss gets cycles, suddenly finds nobody active on his or her table and resets. Clients gets their cycles, and suddenly our 24% boss has 100% HP again. YAY!

We had this happen 4 times on one boss before we went a miraculous 9 minutes without your hardware postponing our activity.

The problem is, I know this is a resource/priority issue, because I manage systems that host 750k-2million concurrent users. And I know it's tricky, so I'm not mad like my friends. I know all of the things you guys could possibly say in response to this issue because I've issued the same responses-- hell, I can't tell you how many canned responses I've had people record into switcher messaging systems. Sometimes the best you can say to the end user is, 'we're aware of the issue and workin on a resolution' even if corporate is only focused on one particular project.

SO! If you are hosting /all/ instanced activity on the user's home server, and I would guess you are, please stop giving the spaceface instances all of the damned cycles during heavy load. Either host them on another resource, find a better way to spread them across the VMs, or give them the same priority the rest of us get. There are about 3 different different ways I deal with this kind of thing when I have a new product that upstairs wants to shine, and I'm sure you guys know more about the specifics of making it happen than I ever will.

Just let it be known, it's breaking encounters. And we know why. If those that decided on this course of action discard our frustration, remind them:

Your company is paid to provide a persistent service. If Verizon's HD network turned my TV into frozen, pixelated garbage for 15-30 seconds about every 15 minutes, do you think I (our they) would accept that amount of loss? Even if it only occurred on 60% of the channels?

Pete looks great, but Paul and Mary are getting marginalized.

Sincerely,

Pete
Honestly,

You would have wiped anyway

Truthfully,
Hatstandard

bdatt's Avatar


bdatt
12.06.2013 , 01:21 PM | #3
I've had terrible lag all week on JC, including SnV last night. At one point it got up to 130 seconds, yes seconds, and still hadn't DC'd me. Power cycled the router and modem and all seemed fixed.

Edit: Nevermind, it's still terrible. This is becoming unplayable.
Kristioli | Jedi Shadow | Kinetic Combat
Pillars of Ashla

ItachiZaku's Avatar


ItachiZaku
12.09.2013 , 10:30 AM | #4
I had a couple huge lap spikes in level 55 warzone (the one with the 3 bunker nodes)

Also, in the level 50 HM FP, the boss fight with the big dude that'll go King Kong on you and kick you all in the air, he was causing me to lag a couple times.

Not sure if that helps, but I've been otherwise pretty fine.
"Be careful Preston. You're treading on my dreams. "

goobot's Avatar


goobot
12.09.2013 , 03:57 PM | #5
This would probably get a response from the devs if you put it in the customer service forum