[Resolved] Independence Server Was Down

[TeeeBOMB 668]

Ok, Indy server, maybe? Just went down. About 18:40 EST on 24 Oct 2019.

 

Both my main and my alt got kicked from Indy server. Apparently many others on Indy got kicked. It later let me and my alt log back on.

 

So was a temp lag/boot. When I logged back on, I was still riding my horse.

 

TeeeBOMB

[Retrograde 12,013]

Yeah was a network issue, apologies. AWS can't come soon enough! 

[Sklo:D 2,210]

7 hours ago, Retrograde said:

Yeah was a network issue, apologies. AWS can't come soon enough! 

 

Selling AWS as the universal problem solver is a bad idea. Cloud fixes networking issues for sure while it will bring new issues in, which aren't present on root machines.

There is not something like 100% uptime in the IT. Hyping people for something which will not be the case will be an additional disappointment. There is a reason why netflix and Co are using thousands of micro services in the cloud instead of one big application in the cloud and this not only for scalability reasons.

[Ekcin 2,417]

Maybe I read a different Retrograde response. In than, he did not do anything like

Quote

Selling AWS as the universal problem solver

Instead, he confirmed that this was a networking issue, and you stated yourself

Quote

Cloud fixes networking issues for sure

That's it. Yes, there may be and will be new cloud related issues, and I am not really in love with clouds much less with Amazon in particular. But from all I know it may be a viable and affordable solution for GC/CC, with a number of advantages over the recent hoster.

[solmark 546]

Good old @Sklo:D , keeping it real :)  Surely even you must agree that in general AWS is an improvement on what we have, right?

[Sklo:D 2,210]

1 hour ago, solmark said:

Surely even you must agree that in general AWS is an improvement on what we have, right?

 

Mhm, I wouldn't be completly sure on that. AWS sounds very fancy, but in the end there is not much experience with running huge applications like Wurm in the cloud. The servers I am using for a Wurm project I work on started in the cloud and have been moved away from the cloud, as there were too many downsides, like freezes when VMs were moved between hardware components without warning, bad I/O performance and in general a lot less performance than a root server could offer, at more costs. Cloud just didn't seem to be a good idea for Wurm back in the days, we will see how AWS turns out could be a success but also could go horrible wrong. The fact that the AWS move already takes more than 6 months shows that this isn't such an "easy 100% sure it will work out" solution as advertised and that there are problems. Promoting AWS to be the hero to safe Wurm with 100% uptime guaranteed is wrong, while AWS server have great SLA levels other problems need to be taken into account as mentioned.