What takes so long?

I’d have more sympathy for them if they weren’t a multi-million dollar company.

1 Like

There is a difference between an outage, and planned 8 hour maintenance.

Multi-million dollar company, which has clearly invested next to nothing into their software development lifecycles. I’ve seen small mom-and-pop companies invest more into their Software release cycle then these guys. I guess that’s my frustration.

3 Likes

First it was 330pm now 6pm. I was ready to do my reputations grind at 9am. I’ve only been on for about the last 2 months after a 5 year hiatus. Now at 6pm i’m done for the day. Hope it’s up tomorrow.

Yup. They even downloaded content in advance and still no one seems to execute digital tasks with finesse astounds me.

2 Likes

Apples and thumb tacks.

Your comparisons are not even remotely close.

2 Likes

If you have any more wisdom to share in response to me which is unrelated to my post and obvious in nature, I do hope you share it.

I can’t figure out what you’re trying to say.

Because it’s a vastly complicated process for them, over an enormous volume of data and infrastructure.

That’s why it takes so long.

It’s very difficult to grasp the complexity of their system. We see this all the time after some patch or update when the most trivial thing breaks. We’ve seen messages like “fixed X” and, suddenly, Y is having the oddest of bugs. How did X affect Y? How was that even possible? Who knows?

What was it, in Bastion, or the taxi in general where it was dropping folks trhough the world and sending everyone to WESTFALL(!?!?!)?

I don’t envy anyone who makes wide spread changes to a 20 year old code AND the associated ENORMOUS databases.

Mr. 24x7 heart monitoring is a far cry from the ball of mud these guys are dealing with.

We won’t go into the smoke testing of such a system.

Finally, this isn’t the space shuttle. It’s also not heart monitoring, nor is it the NASDAQ. If it’s down a bit longer, all it costs is money, their money. Money in labor, money in resources, potential lost revenue.

Y’all think they finished coding and tuning this thing a month ago. As always, they were likely changing things to the last minute, because it’s always the last minute, they’re always deploying, they can always “fix it tomorrow”.

Why does it take so long? Because it’s complicated. It’s always been complicated. Despite their best efforts over time to engineer and make it less complicated, it’s always a losing race.

Yet, YET! It’s always getting better. Compared to earlier efforts, recent launches have been almost flawless in the big picture.

This is a big “patch”. This IS the “game”. This IS DF, it just doesn’t have all the knobs turned on. This is the vast majority of the work necessary to get the game launched in November. It’s because of days like today that will make game launch that much smoother, when folks are going to be REALLY eager to get in and play.

6 Likes

99.9% of the people that comment in these threads have zero clue what they are talking about when it comes to why blizzard does what they do and how bad and ancient their deployment practices are. Blizzard probably doesn’t even employ site reliability engineers or architects that have a clue how to resolve the massive amounts of deployment problems they have and how their entire Q/A-Production workflow is fubarbed.

Your preaching to a bunch of plebs that think they are engineers cause they run a wordpress or shopify page. They don’t have the mental capacity to understand concepts like Infrastructure as a service using Terraform, or configuration management with puppet or ansible. How you can keep an entire system online using Microservice based architecture with Nomad or K8.

As far as we know Blizzard is still running a giant monolithic codebase on some outdated and overworked FX2 servers that should have been disposed back in MOP.

Now back to the plebs ranting and raving thinking they know something.

So do I and I know I know very little about “just pushing an update” to something as complex and large as World of Warcraft. But I can respect what they are undertaking. You must realize with your vast knowledge its just not a few updates. The core expansion is being added in waves, as well as the core engine. Not to mention back end hardware upgrades. So please, educate us.

1 Like

Espresso machine is down. That means they had to doordash coffee and donuts at 7 am. Since they are in California all the Doordash people are using bicycles to save gas money. They are still waiting on their coffee and donuts.

3 Likes

Delays are pretty much always caused by them running into major issues when doing their internal stability tests after the updates are complete. Sometimes things just don’t work as expected when pushed to the live servers even though they were fine on the PTR servers.

1 Like

Just to think. They tested all this to work out bugs and problems, to prevent this and here we are yet again, no surprise… just wait until folks get in and the instance not found or server restarts go lol frustrating, yes surprise no

1 Like

i completely agree…my frustration is that, as someone else has pointed out, they have been doing this for 17 years, so we should be used to it by now. the counter point to that is: Blizzard has been doing this for 17 years, so they should be better at it by now!

the fact that they have huge unveiling announcements that they then fail to deliver on is just another example of poor customer service from a billion dollar company.

1 Like

At least you put some effort into this reply. /highfive

Okay, I’ll be more verbose…

You’re saying that big software companies (the ones I mentioned) DO have downtime in the form of outages. I agree. I’m saying the difference is that while those companies have unplanned outages, you’re comparing them to Blizzard’s planned downtime. So you’re comparing Apples to Oranges.

“Facebook goes down sometimes, so it’s okay that Blizzard plans to take 8 hours to release their software update.” - That doesn’t really resonate …

If they’re unplanned delays, they why did they plan for eight hours of downtime. Good project managers will plan for some “cushion” time. If they know they can get it done in 3 hours, they say it’ll be up in 5. But these guys are planning for an entire day to update software, and I can’t for the life of me figure out why.

1 Like

Oo we found the life of the party here folks.

1 Like