Queues, in my Sharding?

WoW Classic WoW Classic General Discussion
1

This is NOT a stealth argument to expand sharding (we shouldn’t), just some thoughts on how it will interact with queues and server population.

Will there still be queues during temporary sharding?

If the tourist attrition rate is lower than expected, we could end up with very high pops in the non-sharded zones. This isn’t the end of the world, but it is non-Vanilla-like. I assume there are technical limitations to how many players a non-sharded zone can handle as well.

How do we balance a server’s queue to prevent crashes and extreme overpopulation, while still allowing enough folks to make their way through the starter zones?

Will these numbers be adjusted over time? e.g.:

First week: 20,000 concurrent cap while 95% of players are in sharded zones rage quitting
Second week: 10,000 concurrent cap while 50% of players are in sharded zones but some are actually progressing into the world
Third week: 3,000 concurrent cap, sharding disabled

In before #nosharding.

2

Interesting.

Yeah what he said :arrow_down:

3

Queues wont bother me one bit.
Neither will unsharded zones being flooded

9 Likes
4

The inclusion of sharding does not remove the usage of a que system.

I fully expect both. Because they serve different purposes.

1 Like
5

Same and same, but there’s still the question of what’s technically possible.

If we (temporarily) end up with 15k players in the non-sharded zones, can the servers handle it?

If queues kick in to prevent that, how badly will queues stack up as the tourists get stuck in them?

6

If part of the point of sharding is for them to control populations so that servers have healthy long term populations after the tourists pack up and leave, they need to control the limits that are there at launch.

That means there should still be queues at times, even with sharding.

Having too many players on a server early on can risk overpopulation after the tourists are gone.

Queues will encourage the server populations to spread out a bit more, while still not outright preventing you from picking a server if you really want to or a friend is on it.

1 Like
7

Yes.

If Blizzard is seeing a higher percentage of new players hitting 10+ and 20+ they will need to add new servers and divert new players to those servers. Waiting until 60 or “post launch” will be a failure condition. Sharding is only there to ensure that the tourists who never get past 10 don’t bloat the server’s population limits during launch.

One idea is Differential caps. The cap in the Starting Zone can be an active count of 10k or 20k, while people leaving those zones will be controlled by a 2.5k cap. If necessary, people leaving the zones could be either disconnected, or given the option to be put in the queue, or return to the starter zone.

It would encourage people who want to have a nicer server experience to reroll on a new less populated server before they get very far into the game, if a given server looks like its going to be always full.

8

Don’t care. I really don’t. I feel like were being nice to the tourists as is even letting sharding be used in 1 to 10 for launch.

1 Like
9

Do you know anything about the networking code, infrastructure, load average/balancing or any other details?

I wonder why Players think they have the answers to technical questions like this.

Blizzard knows we do not want sharding except in extreme cases like the potential for 5,000 people in Northshire.

2 Likes
10

Can you imagine 5,000 people in Teldrasil trying to do that relic quest in the cave?

11

That’s why as a starter zone it gets sharded.

1 Like
12

Well, the chances of getting murdered by a group of Furbolgs just went down at least.

2 Likes
13

Even with sharding, those boxes had like 3-5 minute respawn times lol.

People are going to quit with 20 people standing around all repeatedly clicking it. Sharding isn’t going to solve the slow respawn times.

It’s going to be madness either way. :scream:

1 Like
14

Oh well. Classic isn’t retail

1 Like
15

They have to use a total server cap alongside sharding of starter zones. They’re likely to extend caps at the commencement of the game (maybe 8k?) and bring them down to levels they’re comfortable with later (4k) as the tourists drop off.

16

This is part and parcel to the entire exercise. Sharding lets them “over provision” a realm. But there’s still a finite number of active players per realm. It’s just that this number will be larger with Sharding, than without. So, queues will be a thing.

During the initial flood, some folks will see a queue, and jump to a different server. Others may just quit outright and come back later. Over time, as the populations stabilize, the events of Sharding should naturally reduce to, ideally, zero, while the number of overall realms either remains static (“they had enough in the first place”), or is slowly grown to compensate for population load.

I am not of the opinion that Classic will grow over time. I’m of the opinion that it will surge, taper off, and then settle comfortably in to a slow decline, with ebbs and flows of folks coming over for short stays, say, during lulls in retail (notably) or perhaps other games (Blizzard and non-Blizzard). Even now WoW is a game you can let sit and come back to, but Classic especially so as it won’t change the way Retail changes. Come back in a month, come back in a year, save for the player population, Classic will be the same.

But even a game with the depth of Classic has a finite lifespan. Players will come and go, new players will come and go, but I don’t see a large “Hey Classic is the best!” pulling in throngs of non-players over long term in to the Blizzard/WoW fold. So, in that light, Classic will not grow. It will peak in the first year.

Which is why it’s so important for the number of Realms to be restrained. They need to be sized for Years 2-5, not opening day. Sharding adds that flexibility to let the system handle an early overload.