Here is why I don't believe it's a DDOS attack.

"

Snorkle_uk#0761 wrote:

these threads are so ridiculous

I'm no fan of conspiracy theories and my soul shrugs every time someone calls out "streamer RNG". But this time around, there are so many holes in their "it's ONLY(!) DDoS and there's nothing we can do about it" (which Jonathan said in an interview a while back.

Now it has changed to "but we do want to acknowledge that we do see the issue and are doing everything we can to mitigate this". When other developers across Europe CAN mitigate this, GGG can too - and should've done so a long time ago. Time? Money? Knowledge? I don't know what the issue here is, but it has been going on for half a year.

I don't care if it's DDoS, crappy servers or not enough manpower/competence to handle the situation. When a game is periodically unplayable (which is has been in Europe), they need to do something. Anything.

"

Angrybagel#7431 wrote:

What's ridiculous is white-knighting for a company that has been shown to flat-out lie to their customers on multiple occasions.[/quote]

how cute, picking out the one thing you can attack without writing anything about his post. No comment from you about how ddos works and why anyone would do it?
you even know what ddos is?

"

I'm no fan of conspiracy theories and my soul shrugs every time someone calls out "streamer RNG". But this time around, there are so many holes in their "it's ONLY(!) DDoS and there's nothing we can do about it" (which Jonathan said in an interview a while back.

ever heard the following? "I'm not a "insert evil entitiy" but...
What usually follows after the but just shows that you are actually a "insert evil entitiy"

You can do a lot against ddos, but actually it all amounts to money. If you don't want or can't spend it, its over. You think cloudflare block those tbs for fun? go on, check their gaming company tier. How i love comments concerning networks from people who couldn't differate between an ipv4 and ipv6 header.

"

Game instance crashes are something they can (near) 100% prevent with a proper system design.
And if they would use big cloud computing provider properly (like AWS) they could also for most practical purposes protect themselves from other symptoms (supposedly) caused by DDoS with proper setup and coding.

yeah sure, lets redesign everything fast. Its only a small game. Or maybe you go with "they should ahve planned it that way!" ? Pick your unrealistic poison. And no, it doesn't run on AWS. That numbers of connection would cost a lot, AWS also may not provide their specialised server structure etc. And no, lol there is no proper setup/coding which protects you from ddos. None, never under any cirmcumstances. Given enough ressources you can kill any service. Problematic systems can be killed with low traffic (like that ssh negotiation bug back then for example)

"

tsunamikun#0433 wrote:

"

Game instance crashes are something they can (near) 100% prevent with a proper system design.
And if they would use big cloud computing provider properly (like AWS) they could also for most practical purposes protect themselves from other symptoms (supposedly) caused by DDoS with proper setup and coding.

yeah sure, lets redesign everything fast.

The game is more than 10 years old. They have been a comparatively large and successful company. They had access to Tencent resources for a long time. This means they had plenty of time and resources to do a proper server design. So yes this is on GGG, they failed. (And I didnt say anything about "fast" - you are making this up if you refer to my comment here.)

"

tsunamikun#0433 wrote:

Its only a small game.

No idea what you mean here.

"

tsunamikun#0433 wrote:

Or maybe you go with "they should ahve planned it that way!" ?

Correct, see above.

"

tsunamikun#0433 wrote:

Pick your unrealistic poison.

Yes its unrealistic. But not because its not doable. And not because its too expensive. My guess is its because their leadership/management is lacking.

"

tsunamikun#0433 wrote:

And no, it doesn't run on AWS. That numbers of connection would cost a lot,

You got some actual numbers? ("A lot" reads like "I dont have a proper argument" to me.) They also dont need to generally run everything on AWS. They just need it as a fallback for bad times. This is how proper companies design things nowadays: Make server/cloud providers exchangeable so that you have more options.

"

tsunamikun#0433 wrote:

AWS also may not provide their specialised server structure etc.

See above. AWS runs "everything" nowadays. They had I-dont-know-how-many-exactly years to make proper server structure/code. Of course switching from "spaghetti" code servers to properly coded ones (no idea if they have "spaghetti" code servers or no but lets assume here) is neither easy nor fast. It needs effort and will. (My impression is that due to leadership they effectively dont have these.)

"

tsunamikun#0433 wrote:

And no, lol there is no proper setup/coding which protects you from ddos. None, never under any cirmcumstances. Given enough ressources you can kill any service. Problematic systems can be killed with low traffic (like that ssh negotiation bug back then for example)

Maybe you missed the "for most practical purposes" part. DDoS is a big thing for AWS and they are well prepared. So if users dont mess up their coding/design its a solvable issue. Just read up on what AWS offers in this direction.

"

tsunamikun#0433 wrote:

You can do a lot against ddos, but actually it all amounts to money. If you don't want or can't spend it, its over.

Isn't it funny, though, when Jonathan goes on record saying it's not a matter of money? Where does that fall into your "expert assessment"? Are you wrong, or is he lying?

Naturally in internet forums there is a certain bias towards bs statements and conspiracy theories, however claiming that saying "the ddos attack story is just a narrative" is a farfetched conspiracy theory is just silly and naive. The EU server problem is an ongoing never ending tragedy, maybe it's not just GGG's fault but it is certainly not an ongoing ddos attack problem...

"

Strickl3r#3809 wrote:

The EU server problem is an ongoing never ending tragedy, maybe it's not just GGG's fault but it is certainly not an ongoing ddos attack problem...

One good thing about this case of EU issues is that they cant blame the affected players (routing or ISPs or whatever) anymore as almost all cases result in instance crashes.
Unfortunately for players this made it even worse. I wouldnt have been able to play on EU servers due to constant instance crashes anymore and switched to NA, unfortunately lots of ping there but at least it works much better otherwise, not 100% though.

It's not just happening on EU servers, but on SEA servers as well. The game becomes unplayable during peak hours, and calling it 'frustrating' is an understatement. I’ve experienced sudden crashes and rollbacks in another game I used to play—the culprits were RMTers who purposely crashed the game to duplicate items or roll back bricked ones. That game never recovered and no longer exists.

Well, everything ran fine till the MTX's stuff started going crazy. I think what ever League that put that warp bubble people like to keep going on the way point that is when I noticed a clear server dip in performance.

What really pisses me off is you can't turn off other peoples MTX. So they can annoy the shit out of you, crash your games with all the MTX clutter, and GGG is too worried about making a buck than worried that people are not playing the game due to how annoying the MTX are.

People keep talking about the EU servers as if they were hit alone, sorry. I played on the DC server and it was horrid. We couldn't even get a map done, I was trying to threw the Caverns in Act II and the game kept crashing so much I just finally logged off. The next day it was slightly better, but still it was pretty crap for the entire weekend.

"

Zrevnur#2026 wrote:

No idea what you mean here.

sarcasm doesn't travel well without tone.

"

Correct, see above.

Yes its unrealistic. But not because its not doable. And not because its too expensive. My guess is its because their leadership/management is lacking.

but obviously they didn't. And it has to have been lacking since the beginning, one would think that wasn't an actual service oriented company, but a passion project.

"

("A lot" reads like "I dont have a proper argument" to me.) They also dont need to generally run everything on AWS. They just need it as a fallback for bad times. This is how proper companies design things nowadays: Make server/cloud providers exchangeable so that you have more options.

The problem is, without actually knowing how many number of service instances, connections etc they have, i cannot know the price.
If you want to play around: https://calculator.aws
But the problem is, even there you can only do it for one region, i don't even know how many different server clsuters ggg has globally.

"

See above. AWS runs "everything" nowadays. They had I-dont-know-how-many-exactly years to make proper server structure/code. Of course switching from "spaghetti" code servers to properly coded ones (no idea if they have "spaghetti" code servers or no but lets assume here) is neither easy nor fast. It needs effort and will. (My impression is that due to leadership they effectively dont have these.)

Yes which makes it even worse. They may even have the will, but they would need a lot of manpower and specialists for that, good luck finding those in NZ.

"

Maybe you missed the "for most practical purposes" part. DDoS is a big thing for AWS and they are well prepared. So if users dont mess up their coding/design its a solvable issue. Just read up on what AWS offers in this direction.

Yes there is a practical limit. But not a theoretical one. Which was the thing I wanted to highlight. AWS is well prepared on standard webservices which they know. (and which are mostly the target of DDOS). But depending on the service setup from ggg they may not be appliable. The DDOS didn't target their webservices but specific server instances. Who knows how effective the Aamazon Shield protection would be if you could trigger some complex computation on the server.
And if you look how those dcs currently looked (fine ping with no stutter until massive lag, disconnect and rollback) it would seem the attacker knows how to completely kill the instance server, this is not only a net disconnect.
Again, this would be GGGs fault, but thats not the point. The point on my side is, there is no hidden magic, some poe2 overbleed or anything else here, I can be a ddos, and in their current situation, they may not be able to do anything against it. They cannot migrate their infrastructure in a few days, they cannot redesign their code in a few days, thats it.

"

Isn't it funny, though, when Jonathan goes on record saying it's not a matter of money? Where does that fall into your "expert assessment"? Are you wrong, or is he lying?

Is the sky blue, or do apples grow on trees?

Its a matter of time mate. Time, manpower, knowledge, money. They have 1, maybe 2.

"

tsunamikun#0433 wrote:

sarcasm doesn't travel well without tone.

I didnt use sarcasm in case you are implying that.

"

tsunamikun#0433 wrote:

but obviously they didn't. And it has to have been lacking since the beginning, one would think that wasn't an actual service oriented company, but a passion project.

Yes. And we have had server/GGG side problems for a long time in EU. And early on it may have been fair to call it "passion project". Early on it would have been "unreasonable" to expect them to do better. Still leaves years (at least post Tencent sale) to fix this.

"

tsunamikun#0433 wrote:

Yes which makes it even worse. They may even have the will, but they would need a lot of manpower and specialists for that, good luck finding those in NZ.

If this "Its hard to find people in NZ" is actually an important real thing instead of a PR argument then this is also IMO a symptom of leadership failure. They should make "the switch" then to a company model that doesnt lock them in like that.

"

tsunamikun#0433 wrote:

Yes there is a practical limit. But not a theoretical one. Which was the thing I wanted to highlight. AWS is well prepared on standard webservices which they know. (and which are mostly the target of DDOS). But depending on the service setup from ggg they may not be appliable. The DDOS didn't target their webservices but specific server instances. Who knows how effective the Aamazon Shield protection would be if you could trigger some complex computation on the server.

Thats why I wrote it the way I wrote it: AWS may not automatically protect them, they may have to do something "extra". However AWS gives them the tools for that.

"

tsunamikun#0433 wrote:

And if you look how those dcs currently looked (fine ping with no stutter until massive lag, disconnect and rollback) it would seem the attacker knows how to completely kill the instance server, this is not only a net disconnect.

Yes, this is also why this "DDoS" explanation looks fishy to me. These symptoms look more like someone using in-game things to use a bug on the server instance to crash it. This however is pure speculation. My point was just to counter the "Its DDoS so we cant do anything."

"

tsunamikun#0433 wrote:

Again, this would be GGGs fault, but thats not the point. The point on my side is, there is no hidden magic, some poe2 overbleed or anything else here, I can be a ddos, and in their current situation, they may not be able to do anything against it. They cannot migrate their infrastructure in a few days, they cannot redesign their code in a few days, thats it.

They can do something which works in the future. It may require effort and will.