The Cloud Engineering Thread of Dominance

[Shaggy3.0 1695]

On 12/30/2023 at 4:37 PM, shnsajax said:

I've taken over our Azure space for my department, but really only utilizing Logic Apps, APIM, and Storage Accounts and Tables. We are moving away from Snowflake and into Databricks which I am pretty excited about.

Need to do more indepth training and eventual certifications, but how much are these really worth for Azure? Also working on getting an Adobe Workfront Fusion Cert this spring too. Need to focus more on being less of a Jack of All Ttrades and Master of nothing having to manage everything with all of the cuts we have made the past 18 months. 

 

Azure/AWS certs are worth the effort.  They challenge you to expand cloud-based technical knowledge.  In my experience, decisions from certified leadership carries more weight when decision-making can be hijacked by office politics.

Can you provide high-level insight (trade-offs) on the decision to move from Snowflake? 

[StassneyHorn 6691]

Any of you that work remote ever try taking on a 2nd job? Had another opportunity reach out after I spruced up my LinkedIn. Was salesforce related not cloud “Cloud”

[blacklab 11993]

I haven't but it crossed my mind more than once.

I did have an employee who did. Started missing meetings and wasn't answering slack and ended up getting fired.

[Updawg 2300]

Make sure it’s not getting reported to a service like work number or ones that verify employment. A lot of employers are using those to look for employees doing this

It’s crossed my mind a few times as well but I’m mainly too lazy to

[Rimbo 26409]

On 5/5/2023 at 8:34 PM, StassneyHorn said:

How many of you are in cloud engineering? I’ve decided to skill up and got Adrian Cantrill’s course to help learn AWS. Currently a Salesforce admin, but looking to possibly transition into a more IaaS- heavy career.

Hi

only just now saw this thread

[Ebeye 116]

I do some side consulting, but nothing that is taking a ton of time.  I always do it above board, reporting it to my current employer through their systems.  It has never been an issue as long as you are watching out for conflicts of interest.  

[StassneyHorn 6691]

Those Azure numbers are crazy

https://www.cnbc.com/2024/01/30/microsoft-msft-q2-earnings-report-2024.html#:~:text=Microsoft beat on the top,Blizzard%2C its largest deal ever.

[Shaggy3.0 1695]

Yup.  Azure and GCP feasting on AWS market share.  It was bound to happen. 

I think it will be a race to the bottom---public cloud commoditization.  I see increased number of hybrid cloud projects.  Cloud repatriation for companies burned with high costs due to lack of cloud migration planning, but whatever.  

The work is there and will be with us for a long time.

 

 

[StassneyHorn 6691]

10 hours ago, Shaggy3.0 said:

Yup.  Azure and GCP feasting on AWS market share.  It was bound to happen. 

I think it will be a race to the bottom---public cloud commoditization.  I see increased number of hybrid cloud projects.  Cloud repatriation for companies burned with high costs due to lack of cloud migration planning, but whatever.  

The work is there and will be with us for a long time.

 

 

I’ve noticed ads for positions called Cloud FinOps and it seems entirely like some firms threw their hands in the air after they bought in on the technology but didn’t bother working with auto scaling or storage lifecycle management. Could be wrong/naive about it but it seemed something an “architect” should know.

9 hours ago, BeardIP said:

I think I posted this on the FIRE retire early thread, but there is a whole movement called "Overemployed" where tech/WFH workers are trying to juggle 2, sometimes 3 and even 4, high salaried positions (to make $500k+/year for a few years) and there are sites on reddit and overemployed.com with strategies and how to successfully do it. 

I don't have that grind and personally find it unethical (I can see the argument for it, however, as well), but I have to admit it's pretty clever and a quick way to 4-5x your retirement plans.

Most know that at some point they will be fired or laid off-- the key is to get junior level jobs or jobs you are over qualified for to supplement your primary job.

https://www.businessinsider.com/overemployed-workers-secret-two-multiple-jobs-salaries-tech-tips-2023-11

Yea that’s where I’m at now. My first job out of college in Austin was working for a broker-dealer in Austin and ACH/EFT and all kinds of very easy to understand cashiering/money movement transactions across accounts. Saw a couple positions at fintech firms paying minimum 70k for that shit when I was making 32k and hating every bit of being in the real world. The fact I could supplement my job with this stuff I could do sleeping is an enticing choice.

[StassneyHorn 6691]

I guess in order to get some hot takes, is a place having a Cloud FinOps role, a sign their architecture sucks, and to stay away? 

[Captainant 27970]

2 hours ago, StassneyHorn said:

I guess in order to get some hot takes, is a place having a Cloud FinOps role, a sign their architecture sucks, and to stay away? 

No, but it's a sign that their governance posture is likely lacking or the org is prioritizing it higher. Most of my finops outcomes that I help my customers achieve are really addressing technical debt that's been outstanding for a LOOOONG time. Stuff like poorly tuned ASG's or "let's just oversize it, the business said it's critical" or an old lift and shift that should be moved to event-driven and serverless. 

By putting a data-driven number behind that opportunity cost of not fixing the tech debt, you can get leaders to give a shit. And then as progress is made, good finops practices will let you claim credit and have the data to PROVE it. My big feather in my cap last year was across the board driving down my customers unit cost of compute by almost 10% just with what amounts to config tuning and smarter replatforming. 

[Shaggy3.0 1695]

On 1/31/2024 at 8:39 PM, StassneyHorn said:

I’ve noticed ads for positions called Cloud FinOps and it seems entirely like some firms threw their hands in the air after they bought in on the technology but didn’t bother working with auto scaling or storage lifecycle management. Could be wrong/naive about it but it seemed something an “architect” should know.

true, an architect should know but it's not that simple.  you have to factor in project deadlines, project team skill levels, technical debt, compliance/legal, data pipelines and potential migrations, multi-cloud dependencies, business ROI and many other things.  throw everything into the mix and you end up with "it depends"-type compromises.  the run-time nature and client traffic for legacy/new solutions will also change your architecture.  lift and shift caused many headaches.  cloud newbs simply migrated on-prem apps to the cloud without full, well-thought out assessments.  there's not enough talent out there to make the hard decisions and accept accountability.  that's why high-level FTE's shift this work to consultants : )

 

   

[Shaggy3.0 1695]

On 1/31/2024 at 8:47 PM, StassneyHorn said:

I guess in order to get some hot takes, is a place having a Cloud FinOps role, a sign their architecture sucks, and to stay away? 

it might be a sign of their overall architecture.  that's a possibility.  but IMHO, it's a definite sign of over-specialization within their org.  in the "old days", high cost and avoiding unnecessary spend was just something you did.  i mean, part of management was to ensure cost optimization for the business.  with cloud, you have devOps and finOps--which means you spend lots of $$$$ on people terraforming all day with the pseudo bean counter oversight to notify on cost over runs.  i think that's really unnecessary but it makes the business "feel" like they're doing something to address high cloud costs.  and honestly, devops and finops can and will be owned by AI in the next few years.  companies with over-specialization tend to over-hire.  this results in layoffs--and i think we're seeing that process take place right now.  

 

 

[BearSchlong 8135]

If you could work for any one cloud/software/services company, who would it be?

I’m thinking sales or customer service.

[MinerProphet 71]

On 8/30/2024 at 6:54 PM, BearSchlong said:

If you could work for any one cloud/software/services company, who would it be?

I’m thinking sales or customer service.

What did you figure out on this question? 

[StassneyHorn 6691]

Just read that AWS announced revenues were up 19% YOY, meanwhile 500+ employees wrote to CEO against the RTO plan. Tough sell unless they want to shed workers. 

"I think we could've been up 25%"

[BearSchlong 8135]

What did you figure out on this question? 

No answer yet. Just waiting on my Q4 POs.

[MonkeyDoughnut 10439]

2 hours ago, StassneyHorn said:

Just read that AWS announced revenues were up 19% YOY, meanwhile 500+ employees wrote to CEO against the RTO plan. Tough sell unless they want to shed workers. 

"I think we could've been up 25%"

Pretty sure reason #1 is they want to shed workers, especially ones they don't have to compensate as they leave.

[immamac 43552]

I'm convinced no one fucking does anything anymore. 

I'm building a company from scratch and we have gotten more done than everyone else in this industry in the last 4 months that we are to the point where we are so far ahead no one will catch us. 

Big tech can't even spend there way to solving problems anymore. 

[Captainant 27970]

2 hours ago, immamac said:

I'm convinced no one fucking does anything anymore. 

I'm building a company from scratch and we have gotten more done than everyone else in this industry in the last 4 months that we are to the point where we are so far ahead no one will catch us. 

Big tech can't even spend there way to solving problems anymore. 

It's the problem with every sufficiently large organization. As an org gets bigger, there's more middle managers that are required to get aligned to actually move ahead with a project and navigate all the corporate red tape. And then when it does happen, there's so many cooks in the kitchen that things get derailed quickly. 

Big tech often has a hard time even correctly enunciating the problem, much less elegantly solving it. The number of projects, initiatives, and plans that I've seen my org and my customers orgs move mountains to realize, only to ultimately find it does not accomplish what they thought it would, is wild.

It's like every year every F500 just careens from one directional extreme to another, dancing to whatever tune is most trendy to investors rather than leaning on their own judgement and expertise 

Edited November 1 by Captainant

[immamac 43552]

2 minutes ago, Captainant said:

It's the problem with every sufficiently large organization. As an org gets bigger, there's more middle managers that are required to get aligned to actually move ahead with a project and navigate all the corporate red tape. And then when it does happen, there's so many cooks in the kitchen that things get derailed quickly. 

Big tech often has a hard time even correctly enunciating the problem, much less elegantly solving it. The number of projects, initiatives, and plans that I've seen my org and my customers orgs move mountains to realize, only to ultimately find it does not accomplish what they thought it would, is wild.

It's like every year every F500 just careens from one directional extreme to another, dancing to whatever tune is most trendy to investors rather than leaning on their own judgement and expertise 

Its all literally lighting money on fire. Its insane. 

[Captainant 27970]

12 minutes ago, immamac said:

Its all literally lighting money on fire. Its insane. 

Just had a crazy meeting today where a customer insisted on setting up their DR such their their on prem would go to our cloud, and then replicate to other cloud (in the same region) to accomplish their "multicloud DR strategy". Nevermind that they have the same directly connected bandwidth to other cloud from their DC and could just push to both targets at the same time from the on prem replication agent. Something about wanting to use all their agent licenses because of some internal initiative on license utilization, nevermind the other costs of the solution like how completely avoidable outbound data transfer is >65% of the total cost. 

But nope, this is they way that the higher ups declared it to be, so that's the way it's gonna be. Just burning money for no good reason aside from a corporate mandate from on high

Edited November 1 by Captainant

[immamac 43552]

Just now, Captainant said:

Just had a crazy meeting today where a customer insisted on setting up their DR such their their on prem would go to our cloud, and then replicate to other cloud (in the same region) to accomplish their "multicloud DR strategy". Nevermind that they have the same directly connected bandwidth to other cloud from their DC and could just push to both sources at the same time from the on prem replication agent. Something about wanting to use all their agent licenses because of some internal initiative on license utilization, nevermind the other costs of the solution like how completely avoidable outbound data transfer is >65% of the total cost. 

But nope, this is they way that the higher ups declared it to be, so that's the way it's gonna be. Just burning money for no good reason aside from a corporate mandate from on high

This has to fucking end. 

[Captainant 27970]

9 minutes ago, immamac said:

This has to fucking end. 

It will continue until morale improves.

[StassneyHorn 6691]

Not a cloud problem but a vent

I’m wasting away in a Salesforce role where I’m mostly explaining healthcare ops errors to departments what product people got wrong and SF developers didn’t know about the end  process. I should be a Salesforce business analyst and get paid more to explain to these people the intricacies, but I can’t do the test developments that always have wrong/incomplete parameters.

[StassneyHorn 6691]

19 minutes ago, Captainant said:

Just had a crazy meeting today where a customer insisted on setting up their DR such their their on prem would go to our cloud, and then replicate to other cloud (in the same region) to accomplish their "multicloud DR strategy". Nevermind that they have the same directly connected bandwidth to other cloud from their DC and could just push to both sources at the same time from the on prem replication agent. Something about wanting to use all their agent licenses because of some internal initiative on license utilization, nevermind the other costs of the solution like how completely avoidable outbound data transfer is >65% of the total cost. 

But nope, this is they way that the higher ups declared it to be, so that's the way it's gonna be. Just burning money for no good reason aside from a corporate mandate from on high

Do they think this is “hybrid” cloud architecture with less risks? How do people sell this?

[Captainant 27970]

23 minutes ago, StassneyHorn said:

Do they think this is “hybrid” cloud architecture with less risks? How do people sell this?

In this case, the customers BOD has decided that multicloud means more resilient. In my experience, that usually means people with a pure business background have been directed as such by some overly expensive consultants, because nobody gets in trouble for following $BIG_CONSULTANT's recommendation. 

It's so ass backwards, even my sales partners don't like it despite the free cushioning on their quotas. But for some reason, leaders of large companies often trust consultants more than cloud vendors when deciding how to best operate in a cloud/hybrid environment 

Edited November 1 by Captainant

[StassneyHorn 6691]

I don’t want to present myself as a cloud expert. Just a Salesforce dude with data analysis skills and AWS knowledge not ready to pay for a cert….. in case you’re hiring.

Are there recruiting level people with no cloud knowledge from your story? I see so many CloudOps and FinOps roles and it’s making me feel like my resume is going into a trash can cause I dont have certain words.

[Zepol87 9263]

Completed the base 3 and have an interview with a big social media company on Monday. Money is great for California but turnover is high. Cloud architecture position and terraform heavy l.

having a network background with datacenter experience helped me 

Edited November 1 by Zepol87

[Ebeye 116]

19 hours ago, StassneyHorn said:

Just read that AWS announced revenues were up 19% YOY, meanwhile 500+ employees wrote to CEO against the RTO plan. Tough sell unless they want to shed workers. 

"I think we could've been up 25%"

When the initial news of the 5 days a week of RTO came out, I messaged some of my friends that were still there.  A couple of them were literally turning in their notices that day.  They all found out about the 5 day requirement through the news feeds.  There were no internal announcements.  

[Ebeye 116]

14 hours ago, Captainant said:

In this case, the customers BOD has decided that multicloud means more resilient. In my experience, that usually means people with a pure business background have been directed as such by some overly expensive consultants, because nobody gets in trouble for following $BIG_CONSULTANT's recommendation. 

It's so ass backwards, even my sales partners don't like it despite the free cushioning on their quotas. But for some reason, leaders of large companies often trust consultants more than cloud vendors when deciding how to best operate in a cloud/hybrid environment 

And most of these orgs are not even taking advantage of the internal resiliency options within the specific cloud that they are in (multi-az/region).  "Vendor lock" and "what if Amazon/Microsoft/Google goes down" are arguments that drive me crazy.  There are not a lot of workloads that are so mission critical that they truly need to be multi-cloud.  It sounds great in theory, but like y'all said, it is just a waste of money more often than not.

[BearSchlong 8135]

Completed the base 3 and have an interview with a big social media company on Monday. Money is great for California but turnover is high. Cloud architecture position and terraform heavy l.
having a network background with datacenter experience helped me 

Ooooh just wait til we get Terreform. Rumor has it it'll be in my bag next year.

[boilerhorn 1446]

On 10/31/2024 at 7:58 PM, immamac said:

I'm convinced no one fucking does anything anymore. 

I'm building a company from scratch and we have gotten more done than everyone else in this industry in the last 4 months that we are to the point where we are so far ahead no one will catch us. 

Big tech can't even spend there way to solving problems anymore. 

Late to the party, but this.  Big tech, with some exceptions, does not want to spend, because they really do not know how to spend.  Innovation, nimble architecture and design, small teams, low latency feedback, etc. are anathema to these folks.  Every idea that is reasonably sound is then assessed by the bureaucracy, inevitibly leading to the conclusion that the company will need to invest 8-9 figures, 4 years, and 500 people to get a product.  And that the product cannot interfere with the existing roadmap, etc.  Many really good ideas are dead on the vine.  No one wants to challenge the status quo, etc.  AND - if there is some internal seed funding - the existing players spend a decent amount of their time trying to derail these new programs.  They think it's existential for them, which is mostly false.

@immamac's example of large results in a short time is exactly what I have seen in small teams and small companies in my career.  LOTS of shit can get done in a very short amount of time.  First, the team is small, so the overall communication network is small (fwiw, it is a function of the square of the # of channels - people in this case).  Second, selection approach is rigorous.  You cannot afford to add mediocre people to the fray.  3rd, each person MUST produce at least 1/n of the output for the company - or should have that mindset at a minimum.  In a 10 person company, knowing that you need to produce at least 10% of the ouput is convicting AND motivating.  Now convert that to a 300, 1000, or 10000 person company or team with a shitload of "here's how we've always done it" mindset from middle management, 'cuz that's the only thing they know.

[StassneyHorn 6691]

On 8/30/2024 at 6:54 PM, BearSchlong said:

If you could work for any one cloud/software/services company, who would it be?

I’m thinking sales or customer service.

Salesforce is hiring 1000 AE to sling their Agentforce product. I’m curious if they’ll just hire back all the people they fired at the beginning of last year and get boomerang employees in a minimally lower pay band.

[Vegas64 678]

8 hours ago, StassneyHorn said:

Salesforce is hiring 1000 AE to sling their Agentforce product. I’m curious if they’ll just hire back all the people they fired at the beginning of last year and get boomerang employees in a minimally lower pay band.

Salesforce stock is at an all-time high (up $50 a share this year alone, if memory serves). Is that a coincidence?

[StassneyHorn 6691]

Beats me, I feel like Salesforce Genie was the soft rollout but then become Data Cloud, Salesforce Einstein is pretty much Agentforce cause they never settle on a fucking name. At this point I KNOW they do it intentionally, can’t be a feeling when it always happens.

Edited November 14 by StassneyHorn

[Vegas64 678]

1 hour ago, StassneyHorn said:

Beats me, I feel like Salesforce Genie was the soft rollout but then become Data Cloud, Salesforce Einstein is pretty much Agentforce cause they never settle on a fucking name. At this point I KNOW they do it intentionally, can’t be a feeling when it always happens.

Nah I disagree. This fundamentally feels different. I've been following because I retail invest in CRM and been in the sfdc ecosytem in a previous life and I haven't seen Benioff this excited and invested and passionate in....ever. This feels more akin to Zuck when he went all in on Meta (and I'd even heard rumors from employees that he was contemplating changing the name of the company to agentforce).

[Zepol87 9263]

Lightning is kind of dog shit, just wanted to pop in and say that 

[Vegas64 678]

10 minutes ago, Zepol87 said:

Lightning is kind of dog shit, just wanted to pop in and say that 

Classic for life.

[StassneyHorn 6691]

Alright, own it, who works at Netflix?

[pacman 7426]

On 11/15/2024 at 10:51 PM, StassneyHorn said:

Alright, own it, who works at Netflix?

Haha, it was funny given the flood of "chaos monkey" marketing by Netflix these past few years, crowing about their infrastructure teams. Has anybody seen any technical analysis of what broke down?

 

On the topic of this thread, I did earn the Solution's Architect from AWS recently. Definitely acronym soup fatigue. I come from cybersecurity side, so we will see how useful it is in my current role. Glad to have the general knowledge anyway.

[Captainant 27970]

1 hour ago, pacman said:

Haha, it was funny given the flood of "chaos monkey" marketing by Netflix these past few years, crowing about their infrastructure teams. Has anybody seen any technical analysis of what broke down?

My understanding of the event is that they tried to roll their own CDN (rather than use a partner like Cloudflare or Cloudfront) and it worked

1 hour ago, pacman said:

On the topic of this thread, I did earn the Solution's Architect from AWS recently. Definitely acronym soup fatigue. I come from cybersecurity side, so we will see how useful it is in my current role. Glad to have the general knowledge anyway

Associate or Pro? Yeah the acronym soup is a little much, but it does mean you've got a good handle on the services and how they fit together, which is often in short supply from the security domain in my experience lol. 

In my opinion the value prop of cloud isn't necessarily doing something better than you could DIY, but the opportunity to offload some solutions you just need at a commodity level, like WAFs or object storage or large scale auditable encryption at rest.