RIP Linkshells

Quote:

---

Originally Posted by SannaR

Cause they were somehow supposed to know that they'd accidentally break Linkshells.

---

They ended maintenance 1 hour earlier, i expected them to you know actually test basic functionality

Quote:

---

Originally Posted by Avoidy

Man I was chill about this from page 1 but even then it's not hard to see why people would panic about this when they log in and every linkshell is purged with no word from square for hours after the fact. It's a real easy mode lukewarm take to come in after things are announced and on their way to a resolution to pretend like you knew things would turn out fine. Especially rich coming from someone who wasn't even on the affected DC.

---

Even if it happend on my dc my take would be the same. :D They said servers were being upgraded and data transfered its not issue free process not my fault believe their own lies.

Quote:

---

Originally Posted by NekoMataMata

I don't know how they managed this. They had a 24 hour maintenance for what seemed like very little changes, and instead of everything going super smoothly, this happened instead?

I'm seriously trying to be fair. I know there's some people that just want to rip on the game just to rip on it but how do you even defend this?

---

I will admit that I'm more impressed by this mishap than anything else. Things are always going to go wrong in some way, but they did have a full day, so...I guess for SE, that doesn't mean fewer potential problems, it just means they have a full day to mess it up in ways maybe even they didn't know where possible.

Quote:

---

Originally Posted by NekoMataMata

I don't know how they managed this. They had a 24 hour maintenance for what seemed like very little changes, and instead of everything going super smoothly, this happened instead?

I'm seriously trying to be fair. I know there's some people that just want to rip on the game just to rip on it but how do you even defend this?

---

Hardware replacement is not a small change. That they only had linkshells break on one DC is a pretty successful implementation for something like this.

Quote:

---

Originally Posted by LaylaTsarra

Major server switch over and if the only bug we get is Crystal DC linkshells not working I'd say that's pretty amazing. I suspect there will be other bugs and hotfixes incoming.

---

This. I'm fine if people don't have infrastructure of software engineering experience but when we who do have that experience say "It's not that simple" then we shouldn't be called white knights for it. It's not that simple, people.

Quote:

---

Originally Posted by Saraide

Is it just me or do long maints feel significantly jankier than short ones.

---

Makes sense. Shorter maintenance is for smaller, quicker changes that are less likely to have unexpected "side effects." Longer maintenance isn't just more time to do the same thing...it means the maintenance itself that is happening is significantly more involved. The more complicated it is, the more likely something goes wrong and needs an extra fix.

Quote:

---

I don't know how they managed this. They had a 24 hour maintenance for what seemed like very little changes, and instead of everything going super smoothly, this happened instead?

I'm seriously trying to be fair. I know there's some people that just want to rip on the game just to rip on it but how do you even defend this?

---

Doing a massive hardware transfer is the polar opposite of "very little change." It's a huge undertaking from an IT perspective, and extremely easy for anyone in the field to "defend" it.

Quote:

---

Originally Posted by TaleraRistain

Hardware replacement is not a small change. That they only had linkshells break on one DC is a pretty successful implementation for something like this.



This. I'm fine if people don't have infrastructure of software engineering experience but when we who do have that experience say "It's not that simple" then we shouldn't be called white knights for it. It's not that simple, people.

---

Oh yeah, it was no doubt a lot of work. I'm not doubting the effort involved at all. I just find it amusing that they had a full day to do it and they still broke something. :p

I'm surprised some people went alil "sparky" with their interpretation of others intellect out of bookshelves. [Eh, ah well, Reddit has a way of screenshotting us now to mock us I saw 20 mins ago]

And this wasn't only just Hardware Replacement and all that entails.
But other things that tie into it, as other Users have stated before. <and I tried to hint before hand without being blunt>

Hence why I'd be curious to know the broadworkings which, like I hinted before, someone that would know, was unlikely to be here at this exact time.

Either way, did things actually get back to normal now.
As some of us have returned from their slumber?

Quote:

---

Originally Posted by GTK0HLK

[Eh, ah well, Reddit has a way of screenshotting us now to mock us I saw 20 mins ago]

---

Reddit is far worse all around than we are, so their opinions don't hold a lot of weight.

Quote:

---

Originally Posted by Vahlnir

Oh yeah, it was no doubt a lot of work. I'm not doubting the effort involved at all. I just find it amusing that they had a full day to do it and they still broke something. :p

---

More time just means more things that can break. You'd be surprised how often bugs come up in production data centers.

Quote:

---

Originally Posted by Vahlnir

Oh yeah, it was no doubt a lot of work. I'm not doubting the effort involved at all. I just find it amusing that they had a full day to do it and they still broke something. :p

---

They had a full day to do the implementation but software and the hardware it works with sometimes does unexpected things, and sometimes it's not even something you can catch in testing since it doesn't manifest until all the pieces are chugging along together in a production environment. All said, this is a pretty minor bug for the changes that had to be made. Though I imagine there are probably some more that will come up.

Quote:

---

Originally Posted by Vahlnir

Reddit is far worse all around than we are, so their opinions don't hold a lot of weight.

---

Oh I was just joking that some random persons opinion mattered so much, they had to put a user on blast regardless of if they were part of it or not.
<which could constitute harassment in some cases. but they'll use the hidden name tactic. despite the fact the user was forced to change to an alt to hide face.>

but this is off topic.
[besides I find it easier to judge in case by case. if someone isn't going to be fair, they shoot themselves in the foot. not the other way around, as funny they think their troll is, which it isnt.]

Quote:

---

Originally Posted by TaleraRistain

They had a full day to do the implementation but software and the hardware it works with sometimes does unexpected things, and sometimes it's not even something you can catch in testing since it doesn't manifest until all the pieces are chugging along together in a production environment. All said, this is a pretty minor bug for the changes that had to be made. Though I imagine there are probably some more than will come up.

---

Sadly we suffer from too many armchair devs in here that can't give good input other than small indie company please understand where it doesn't matter how big you are...stuff can happen regardless. Heck apparently Ashes of Creation is going to launch with no issues....yeah no MMO ever has done that.

Quote:

---

Originally Posted by CStrife912

Sadly we suffer from too many armchair devs in here that can't give good input other than small indie company please understand where it doesn't matter how big you are...stuff can happen regardless. Heck apparently Ashes of Creation is going to launch with no issues....yeah no MMO ever has done that.

---

Sometimes when you're bigger it's even more of a problem because either more moving pieces or more chefs in the kitchen. My company is on the smaller side so there's a smaller pool that has to communicate and coordinate together. We use outside consultants for some things too (which is another possibility for SE) and that just brings in even more headaches.

Quote:

---

Originally Posted by Shinkuno

They ended maintenance 1 hour earlier, i expected them to you know actually test basic functionality

---

Remember when the housing lottery 0 bug presented itself only on the live hardware and not in their testing environment?

You can do all the internal testing in the world you like, some things only manifest on the live product, this tale ain't anything new in the world, bugs/issues appearing in live versions of software regardless of how much testing is done internally happens all the time across countless companies across the entire world.

Quote:

---

Originally Posted by CStrife912

Sadly we suffer from too many armchair devs in here.

---

As a former software developer in the game industry and current network engineer, the armchair developers make for some good reading for when I need a laugh though. The people who say stuff like "it's simple" are the best ones.

Quote:

---

Originally Posted by Vahlnir

I will admit that I'm more impressed by this mishap than anything else. Things are always going to go wrong in some way, but they did have a full day, so...I guess for SE, that doesn't mean fewer potential problems, it just means they have a full day to mess it up in ways maybe even they didn't know where possible.

---

Considering the full scope of what such a move would entail based on my own experiences, the fact they did all four NA datacenter hardwares in under 24 hours with only CWLS borked on only one DC as the only issue is actually an impressive feat. I'm actually more surprised that there wasn't more issues in the end, Square's got some great engineers working for them that no doubt were working crazy overtime to get this job done.