@factor13co

Boutique cloud consultancy helping you get the most out of AWS, GCP, and other cloud platforms. @ericthewolak@hachyderm.io

 Seattle, WA
hachyderm.io/@ericthewolak
Joined April 2018
  • Tweets200
  • Following285
  • Followers25
2Media
The argument that "we need to use microservices to get hard boundaries between components," is really "we are so undisciplined that we need to add considerable complexity to protect ourselves from ourselves." →
50
240
29
1,591
i think it was the google SRE book where i first heard this rule of thumb -- you can plan to 10x your current state /at best! if you're lucky!/ -- but anything past that that is foolish, pointless conjecture that will bear no resemblance to reality.
You’re unable to view this Post because this account owner limits who can view their Posts. Learn more
1
6
56
A lame service mesh is the worst. If services can't reliably reach each other, the whole dream of microservices falls apart.
Factor13 retweeted
Yishan
@yishan14 Jul 2022
What we did was fail to contain a pandemic when we could have closed borders, done a short, sharp lockdown to isolate and ride out existing infections, vaccinate everyone before variants arose, and then adopt a mask mandate to keep r0 of remaining infections under 1.
12
82
8
722
Factor13 retweeted
Yishan
@yishan14 Jul 2022
In total, you're looking at about 10-15% of your team out sick or compromised. Covid seems to cause brain inflammation, people who are infected (or long-covid) often make errors in their work that they don't even detect.
5
40
6
523
Factor13 retweeted
Yishan
@yishan14 Jul 2022
This means that nearly 5% of the working population is compromised at any one time. With variants evading both vaccines and prior immunity, and no mitigation methods in place, people will continue getting reinfected over and over, so no herd immunity will be reached.
8
80
6
620
AOC is the only leader I've heard articulate any sort of plan. I don't agree with her on everything, but at least she has specific ideas and is willing to try to implement them.
3
1
1
13
I have never in my adult life felt so detached from my party. Biden seems to have no clue what is going on. Every elected leader seems to have one play and one play only - send contextless emails and texts asking for money. WHAT IS THE PLAN?
1
1
1
19
Every time I think about the Democratic party, I feel angry. We are a country in suffering. A majority of us want change, and the Dems seem absolutely unable to capitalize on that. The most enraging thing - I don't even think they're trying.
4
1
28
Is this marketing just AI generated at this point? What is a cleanroom for my lakehouse??
Databricks
@databricks7 Jul 2022
Cleanrooms can transform your security posture & open up use cases. But current solutions have their shortcomings. Until now! Meet data #cleanrooms for the Lakehouse, which enables organizations to share data on any cloud in a privacy-safe way ⬇️ dbricks.co/3QRfVzk
Can we acknowledge how ridiculous it is that startups and even individuals can bankrupt themselves... ... with an AWS bill? I cannot fathom why Amazon does not do anything meaningful to allow setting e.g. limits on accounts. Why do we need to hear stories like this on repeat?
8.31am. I got a call. We ran up a $65,345 AWS bill without realising. I remember the day vividly. I got woken up by my phone ringing. 'Hello?' It was someone from Amazon. Our card on account had been declined. I knew there was money in the bank, so I said I'd check. I asked for clarification on the amount. '$65,345', they said. My heart sank. Stomach turned. I felt a wave of panic flush over my entire body. I didn't know what to do, so I said let me look into it, and hung up. First I called my Co-Founders, told them to sit down, and let them know the situation. Next, I called our head of engineering at the time, and tried to figure what the hell had happened.

ALT 8.31am. I got a call. We ran up a $65,345 AWS bill without realising. I remember the day vividly. I got woken up by my phone ringing. 'Hello?' It was someone from Amazon. Our card on account had been declined. I knew there was money in the bank, so I said I'd check. I asked for clarification on the amount. '$65,345', they said. My heart sank. Stomach turned. I felt a wave of panic flush over my entire body. I didn't know what to do, so I said let me look into it, and hung up. First I called my Co-Founders, told them to sit down, and let them know the situation. Next, I called our head of engineering at the time, and tried to figure what the hell had happened.

257
549
203
3,986
Working at different types of companies - e.g. "traditional" companies, big tech, startups, agencies, finance etc - you realize this: Different software engineering approaches work better/worse for different companies. There's no one silver bullet across the industry.
16
40
7
530
Loyalty is critical in business, but there are different kinds of loyalty. Loyalty to people, loyalty to the business, loyalty to ideas. Know which one you subscribe to, and know what others do. They don’t always mix well.
1
This is spot on and applies to individual contributors too. If you are bringing a very senior engineer who only worked at big shops, they will get confused why they can’t hand off as much as they used to.
Gergely Orosz
@GergelyOrosz31 Mar 2022
Replying to @GergelyOrosz
A major risk at small startups is the culture clash of bringing a manager in who has only worked at the big places from their first job. There are exceptions, but eg someone with 10 years of Microsoft & Google combined often has trouble when hired to lead at a scrappy scaleup.
3
10
1
115
I do sometimes worry, though, that I've gotten too *expensive* for writing a lot of code to be a positive-ROI activity.
Jaana Dogan ヤナ ドガン
@rakyll24 Mar 2022
You are never too senior to write code. I only laugh at people who believe they are too good to write any code as a software engineer.
Leaders with highly scheduled, rigid calendars (“sorry, can’t help now, my next opening is 2 weeks out”) make it hard to operate well in a fast-paced environment because their schedule isn’t nimble enough. They are in essence dealing this week with the priorities of 2 weeks ago.
21
41
6
511
Do it for the 💰
This Post is from a suspended account. Learn more
Every few years I awake from my slumber, look around to see if frontend has become a stable platform, then go back to backend for another few years. These folks are doing great work, but the platform and target are still under so much flux. I miss RoR.
Andrew Lunny@alunny8 Mar 2022
Over the weekend, my team at @stripe converted the company's largest JS codebase from Flow to @typescript. We modified about 3.5 million lines of code, and then hundreds of developers came in Monday morning ready to write TS.
1
Is MapReduce just MPI for Millenials?
1
Microservices are a natural consequence of a DevOps model of on-call. If you’re going to be woken up in the middle of the night because “Calypso is throwing errors” then the scope of the service had better be small enough that one SWE can reasonably debug it.
1
1
1
more replies
If you’re going to have people who mostly write features be on-call for a production system, that system needs to consist mostly of code familiar to the active on-callera. A rotation larger than ~12 falls apart, so a service can only get so big before being unsupportable.
1
If you instead have dedicated on-calllers (e.g. SRE model), they can spend far more time familiarizing themselves with the system, and therefore they can support a much larger system. SWEs don’t play Wheel of Misfortune because they’re too busy building features.
Load more