A History of EVE Database Server Hardware

Yes, Windows Server 2022 for azure, i agree.
Still it’s nice to see what’s currently in use.

So with luck, no more boot.ini overwriting, instead they will be clobbering Grub config files (if somebody is stupid enough to elevate to root, Windows didn’t protect boot.ini at all).

If ever you needed a reason for a read only media protected boot partition, it’s Eve Online.

Migrating to a different database isn’t a small job.

Especially when there’s a bunch of database procedures being used (which, I believe, there are in Eve.)

You also may need to swap out your DBA staff, which is far from a small change.

1 Like

So many flavours of SQL to choose from, dump SQL and use relational calculus.

Just making sure you’re aware:

Pretty much the only thing running in python is the actual game mechanic code. Everything else has been running in something else for quite a long time.

(ESI’s been heavily based in go for a long time now.)

1 Like

Time for that to go too. :smirk:

I hope you realise how much additional work that would introduce to do. Work which would touch on most of the game experience. When you have to reimplement all the game mechanics. All of them.

2 Likes

Well, they’re pretty much gutting it anyway with Quasar? If not then it’s dancing around the elephant in the room at the periphery.

Next blog will focus on the software side of the database config with all that fun stuff.

6 Likes

Our next blog will focus on the software side, but you are right, it is Windows 2019 and it’s Standard.

I don’t think the DC version gives us anything unless its being used as a hypervisor host.

3 Likes

Just the casual 32GB/s connection and 4TB Ram 120TB HDD… Almost as good as my laptop :rofl:

Crazy how much you need to operate the biggest shared excel sheet in the world =^_^=

2 Likes

Yep.

Main use of datacentre is to license whatever host you’re running more than 7 windows server vms on.

1 Like

If only Linux / Bsd had a server/data center capability and could run VMs.

we’re indeed heavily invested in stored procs - not to mention a few enterprise edition features of MSSQL - in particular table partitions for swapping out entire blocks of data for deletes.

Dev Teams are and have been using other database tech for their features for some years now however, so we’re far from ONLY using MS SQL.

There’s at least CosmosDB/PostGres and a few other cloud based managed DB services being used.

Horses for courses and such

1 Like

if you’re running windows VMs, you still need to license them, regardless of the virtualization software in use. By licensing the underlying hardware with datacentre edition (regardless of if it’s actually installed there) you’re covered for an infinite number of windows VMs on that hardware.

Standard edition is 2 VMs per set of licenses. It’s cheaper for like 6 vms, but there’s a switchover after that.

The joys of licensing.

1 Like

So which is more reliable? AWS or Azure?

CosmosDB is Azure, chat is on AWS?

The goal is to reduce dependency on that :smirk:

Using their cloud services and infrastructure is one thing, being trapped into a proprietary non-relocatable stack is another.

And probably why MS won’t do Qt on .Net MAUI due to the Qt commercial dual license (everybody else uses Skia).

Licensing is great when it works for you, not so great when it doesn’t.

I’ve not done any real in-depth checking into which is more stable.

They both seem fairly reliable though and I’d guess instability comes more from us changing things and bugs being fixed/introduced than actual underlining platform issues.

It’s difficult to baseline and compare things over time when the code base changes daily

3 Likes

On the exefile.exe process there is a parameter called /machinehash, what is the parameters (presumabily made up of values identifying my machine) of that value and algorithm used to compute the hash and what is the purpose of it?

It’s a short hash (32-bit) so a weakish algorithm (MD5? Truncated SHA-1?)?

This would qualify as PII wouldn’t it under the GDPR (as far as I can read it is qualifying as PII as pseudonymisation is not the same anonymisation and can be used to identify as per Recital 26 of the GDPR, even if you cannot reverse the hash, you can still attribute it to an individual, even individuals whom have no account with you if using another machine).

From the Privacy Policy - CCP Games

Which data components are making up the machine hash and how would using Wine on Linux affect that?

If we change hardware configuration, so what? That means one component would need to involve the client hash in order to determine unauthorised modifications, so why not just use the client hash? That’s IF the purpose is as per privacy policy statement, to detect unauthorised clients. Knowing the machine configuration says nothing about the legitimacy of the client code.

The keywords here in the CCP declaration of purpose of use are

image

That seems to be overreach and outside the stated declared purpose. The items listed cannot be used to determine legitimacy (approved by CCP) of the running game software. It also states may, however this value is passed in EVERY time the game client is launched. That contradicts itself.

However, it very well can be used to fingerprint and identify an individual (which is not what that privacy policy section is declaring).

As a system administrator of several IBM servers, I’m fascinated by game server architecture and would love to take a walk in CCP’s data centre!

3 Likes