Jump to content

Shit's Borked


RPM

Recommended Posts

2 hours ago, thunderlounge said:

Had the same issue last night on Firefox as well.

 

2 hours ago, Radical Larry said:

It’s so hard to find good help these days.

There was a bug in the caching that wasn't consistently caching the global javascript file correctly in the software. I disabled some caching functionality so that it would work normally again. Sorry I didn't get to it until late I wasn't around a computer all night.

It should be fixed (I haven't been able to reproduce it at all after trying multiple browsers)

Link to comment
Share on other sites

On 2/19/2019 at 6:24 PM, blacklab said:

With haproxy I can redirect http to https and host the certs inside it. I can also redirect surly to www.surly there as well, eliminating the need to do it on each box. I can also send different folders to different boxes.  

 

You can do all of that on the ALB. I'm doing that right now with my current gig. I'd wager a nickel that the ALB is just HAProxy running on an EC2 under the hood, anyway.

 

On 2/19/2019 at 6:24 PM, blacklab said:

aws is great if I'm not paying the bill, but I'm always the one getting the phone call from my boss on why our aws bill is so high and then have to track down what the idiot developers in our sister groups are doing. Like setting up a RDS cluster for a program that runs once a month for about 4 hours. There was no reason not to just have mysql running on the server that ran the script. So $2500 a month for 4 hours of usage and 696 hours of idling. 

Yeah. That's... really not smart of them. Though I wouldn't consider that an AWS problem as much as a "Devs don't consider costs" problem.

 

Quote

I did interview at aws for a technical account manager role, and didn't get it, mainly because I got in an argument with one of the managers that iron is faster than virtual machines. 

OK, now that is some prime-grade idiocy. You shouldn't have even had to provide examples. The only metric by which VMs can be faster than bare iron -- and I emphasize can be -- is time-to-market. But execution time? I mean... if you understand even the minimal basics of how the technology works and what it does... I mean, that's just ... Was this the sort of person who needed other people to tie his shoes for him? My God...

Actually, that's like those guys from Best Buy who try to convince you to spend $150 on a gold-plated HDMI cable with Fuckin' Magical Bit-Enhancing Powers. That's the best comparison I can think of. Because if you have even the foggiest notion of how tech works, bullshit detectors should be going off all over the place.

I'm curious to know what team this guy was on. It might explain a few things.

Link to comment
Share on other sites

I'll check out ALB. I looked for it before and I think it's one of the things we aren't allowed to use. The standard load balancer is a festering hunk of shit with very few options.

Yeah, they are fucking dumb. But they are driven to things like this due to our dumb ass IT group and their stupid rules. The script in question hasn't changed in years. But we have to have a Dev, QA and Production server for it. It took months for me to convince them to allow us to shut down the Dev and QA boxes. The Dev box had not had a user sign in in almost a year. Can we use Aurora which will autoscale and shut down when not in use, saving us thousands of dollars a month? Nope. But fuck them I'm using it anyway. Somehow they forgot to turn my permissions for Aurora off so I'm going to use the shit out of it till someone says something. 

 

The guy I was arguing with is the manager of all the Technical Account Managers based in Dallas. He wouldn't have been my boss, but an equal to the guy I would have reported to. I would have reported to him when my boss was on vacation. 

Him: "it's impossible for bare iron to be faster"
Me: "It's going to be slower as you have the whole hypervisor layer between the OS and the metal which will slow things down"
Him: "no it won't, you don't know what you're talking about"

At that point my head was about to explode and I realized I probably didn't want to work there.

He leads a team of engineers that tell customers how to best use AWS.

 

 

 

 

Link to comment
Share on other sites

21 hours ago, blacklab said:

The guy I was arguing with is the manager of all the Technical Account Managers based in Dallas. He wouldn't have been my boss, but an equal to the guy I would have reported to. I would have reported to him when my boss was on vacation. 

Him: "it's impossible for bare iron to be faster"
 Me: "It's going to be slower as you have the whole hypervisor layer between the OS and the metal which will slow things down"
Him: "no it won't, you don't know what you're talking about"

 At that point my head was about to explode and I realized I probably didn't want to work there.

He leads a team of engineers that tell customers how to best use AWS.

I know exactly who you are talking about lol. I think you got a bad beat on your AWS interview, but don't let one guy's being an asshole color your entire view of cloud architecture in general.

I'd bet the point he was trying to make is that if you're running bare metal in a datacenter, you'll never have access to as much horsepower and aggregate processing speed as you would running in a cloud environment, without spending a shitload. Metal is absolutely the champ at fixed-size, non-scalable workloads. And if you have a stack architected to best take advantage of a physical infra architecture, then no shit it isn't going to translate as effectively to the cloud as-is.

The whole idea of cloud architectures is to decouple your processes and services from eachother so it's more event-based and less stateful, so that you can work on it all asynchronously and scale specific portions of your stack in response to what needs to be working more quickly, rather than just moving a big hefty monolith onto a bigger box.

Link to comment
Share on other sites

yep, I agree with what you're saying, and that may have been what he was arguing. If your app/site has a chance of going from 100 users to a million overnight then cloud blows metal away. If you're adding 1-2 customers a day and they are all doing huge queries on a large database metal will always win.

I did get an email in December saying "hey we fucked up and we'd like you to come in because we think we want to hire you now". I did not reply. I did tell my boss about it and I got promoted a week later.

Still don't think I could work for that guy. I've had a dozen or so interviews since I got out of the navy in 96. That was only the second time I didn't get a job offer. I remember thinking after I got rejected that at the rate they are growing they are going to have to hire thousands of TAMs over the next few years. There is no way they are going to be able to do that if they are so picky they won't hire me. Not a whole lot of guys with 20 years of hosting experience who can actually hold a conversation with a customer.

 

  • Like 1
Link to comment
Share on other sites

On 2/25/2019 at 5:13 PM, blacklab said:

The standard load balancer is a festering hunk of shit with very few options.

It is known.

It's not just the lack of options. It's that if you're doing multi-AZ load balancing, you have to have an equal number of instances in each AZ, or else it will send too few or too many requests to one AZ or the other.

ELB "Classic" has basically remained unchanged since AWS was first launched. EC2 has had a shit-ton of upgrades. S3 has had some massive upgrades. ELB hasn't changed for shit.

Quote

Him: "it's impossible for bare iron to be faster"
Me: "It's going to be slower as you have the whole hypervisor layer between the OS and the metal which will slow things down"
Him: "no it won't, you don't know what you're talking about"

At that point my head was about to explode and I realized I probably didn't want to work there.

He leads a team of engineers that tell customers how to best use AWS.

 

My head is exploding just looking at that.

And I'm familiar with those teams. They're the people who recommended we partner with Rackspace in designing our AWS setup. I'm like, "We just left Rackspace because of their incompetence." And they're as lost trying to keep up with all of the new AWS products as anyone, so it's the blind leading the blind.

Link to comment
Share on other sites

On 2/25/2019 at 8:01 PM, immamac said:

Let me know if this is still an issue. I did some maintenance today, but nothing that should effect sign ins

Signing in is borked in IOS again. It was fixed, but now the sign-in menu on mobile is non-responsive. I have to click into a thread to sign in. 

Link to comment
Share on other sites

20 hours ago, Apep said:

It is still unresponsive if I'm logged out. If I log in on mobile via the button at the bottom of a thread, the menu starts working if I tap on it.

IOS 12.14, iPhone 7+, Private mode enabled on Safari.

Clearly, you need to quit logging out of Surly.  Root Cause identified.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.



×
×
  • Create New...