Hacker News new | past | comments | ask | show | jobs | submit login

"move to the cloud" is a top strategic objective in my org

I have explained why it should not be, but Mgmt does not care.

Everyone's doing it. We NEED to do it.

I believe we are in a mix of FOMO, resume-driven development and empire-building.

Ok, I can tell that you want me to expand. Here goes:

Cloud may offer:

1) improved scalability (both horizontal and vertical)

2) improved availability

3) reduced cost

None of these are guaranteed and will require much expertise in both initial choice-making and continuing execution.

We don't really have that expertise.

Do you?




> 3) reduced cost

From what I've always heard, of all the benefits the cloud has, reduced cost is not one of them unless your compute need is exceptionally small (ie, you only need a few t3.* instances).


That’s because nobody who cut costs feels compelled to go out and proselytize about it. At the start of the pandemic I helped dozens of customers cut their spending to $XX pm as they were inherently crippled during COVID (think travel, hospitality). My management team at AWS supported this effort. It was actually one of our strategic goals.

Anti cloud zealots are having their time just like pro cloud zealots did. You have to understand your workloads and cloud offerings to see if it’s right for you.


As one of the loud "anti-cloud zealots" out there, I agree with you that this is actually a workload-dependent decision (or a luxury product that you can buy to show that your company is rich and modern). I think a lot of big companies are going to be moving toward "hybrid cloud" in the near future if they haven't already: maintaining a few racks (or an entire DC) to handle their base load and then doing bursts of heavy computing in the cloud.

The days of the full cloud lift-and-shift are going away, and we're getting to the point where people make that decision rationally.

As an aside, I actually believe in the cloud enough that I am working on a startup for a piece of cloud infrastructure.


"You have to understand your workloads and cloud offerings to see if it’s right for you."

Yes

This is hard to do.


> None of these are guaranteed and will require much expertise in both initial choice-making and continuing execution.

I feel confident saying that the government does not have this expertise and never will, so this just seems like yet another taxpayer-funded boondoggle (aka, the status quo).


how does improved scalability, improved availability, and reduced cost help the patent examiners make more reliable decisions?

It doesn't, that's the point, you're adding to the same noise everyone else is.


You're right. It's just a random side convo that came out of the main thread.


Reduced cost.

Yeah, we’ll see there. This is the federal government, not a startup.


Our company matches what you describe to a T. We have some small web servers/services, but none under any significant load. No need to scale dynamically, yet we use Docker (with Swarm) on prem for like 3 worker nodes. So much complexity added because of FOMO and resume-padding.

And then the key decision-maker says if we don't do cloud architected systems, we'll have trouble recruiting developers with suitable skills...




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: