dastbe

22 ms·

181.
by dastbe 5y ago
no, because most applications don't have an online dependency for creating roles and modifying policies. what they do typically have an online dependency for is provisioning credentials from those roles, which is architected to be regi
182.
by dastbe 5y ago
(I used to work at aws) The control API (i.e. adding/removing roles, modifying policies, etc.) is available out of us-east-1. However, the bits of IAM that relate to distributing credentials to instances/tasks/lambdas and STS
183.
by dastbe 5y ago
my take is that the overwhelming majority of services insufficiently invest in making testing easy. the services that need to grow fast due to customer demand skip the tests while the services that aren't going much of anywhere spend w
184.
by dastbe 5y ago
> But, the DynamoDB team is running a mostly-regional service. this is both more and less true than you might think. for most regional endpoints teams leverage load balancers that are scoped zonally, such that ip0 will point at instances
185.
by dastbe 5y ago
that’s an update for amazon connect, which is a customer support related service.
186.
by dastbe 5y ago
aws recently launched cloud control, which more or less decouples the individual resource management aspect of cloudformation from everything else. providers like terraform/pulumi can use it in lieu of manual integrations with every aw
187.
Xen-on-Nitro: AWS Nitro for Legacy Instances(perspectives.mvdirona.com)
1 points by dastbe 5y ago|0 comments
188.
by dastbe 5y ago
you make it an optional field. it’s converting an optional field to required that is effectively impossible
189.
by dastbe 5y ago
i think they are higher than you expect, because usually what causes the bug to be known is a worsening state of the system that makes the bug more likely to be hit. i would ask how the engineer found the race condition, and whether that do
190.
by dastbe 5y ago
not the op but aws made the same determination. the tl;dr is that the surface area of containerization leads to an unacceptable risk of privilege escalation.
191.
by dastbe 5y ago
data interchange formats try to encode as little backwards incompatible information as possible. in this case, it would be the restriction that something is a sum type when it could have multiple fields set in the future. another example is
192.
by dastbe 5y ago
imo at least with dynamodb, its pretty straightforward what any given call is capable of doing, and you can get back capacity usage on every request to track. with these mysql/postgres compatible apis the actual query is abstracted as
193.
by dastbe 5y ago
your comment surprises me, because this kind of development is explicitly trying to avoid bifurcating the language. i would argue that the async io/futures libraries that exist in java are the bifurcation because programming with them
194.
by dastbe 5y ago
article from nytimes with more context and uber's response: https://archive.md/AHBG4 not to pick a side, but doesn't the crux of the lawsuit fail if uber is now waiving wait fees for people who have a disability?
195.
by dastbe 5y ago
i’ve seen at least two cloud services using akka for a component that required long-lived stateful connection management. while i think both ended up being successful because of akka, it required retrofitting a lot of libraries that already
196.
by dastbe 5y ago
i’m very interested to hear more! it’s hard to speculate without more details, but based off the bits here i’m suspecting that their load balancing/service discovery system hit an issue that resulted in a death spiral. often times a lo
197.
by dastbe 5y ago
the "nice" way of doing this would would be to create a list of your stringified arguments, mapped urlencoding over them, and then join them with the parameter separator. this ends up being resilient to someone adding something th
198.
by dastbe 5y ago
i do get where they're coming from. right now we rely on dang and other mods (do they even exist?) doing the right thing in terms of making benign and beneficial changes to the linked story and being visible about making those changes.
199.
by dastbe 5y ago
you can see a similar lack of understanding when they use a browser-based test of http/1.1 vs http/2 to justify the improved performance of http/2 for service-to-service communication. what that test is really demonstrating i
200.
by dastbe 5y ago
Is this because of how different col is between SF and Atlanta? Because I'm looking at https://www.levels.fyi/company/Mailchimp/salaries/Software-E... and the compensation doesn't seem outrageous.
201.
by dastbe 5y ago
why do you think the locals would yell piñata
202.
by dastbe 5y ago
so caveat this with the fact that i used to work at aws, but one thing aws is really good at is listening to customers and reacting. a few years back aws messaging was very much so single-cloud and no on-prem. but after listening to custome
203.
by dastbe 5y ago
also, I didn't downvote you but your comment was uncharitable to the people building eks anywhere and was basically conspiracy theory. iirc, it was something to the effect of this product making people hate on-prem and ultimately movin
204.
by dastbe 5y ago
a slightly different way to view it is that there is a class of customers that needs to do something re: kubernetes on-prem and in-cloud. in an ideal world they would like to have someone solve all of those use cases for them, preferably in
205.
by dastbe 5y ago
(I used to work at aws) yes! specifically, quite a number of teams use fargate and its one of the 2-3 "obvious" ways to build services.
206.
by dastbe 5y ago
you could make sure 90% of the money spent on high speed rail went to americans. that’s orthogonal to what the outcome of paying those americans is.
207.
by dastbe 5y ago
every statistic is a summary and a lie. p50/p99 metrics are good in the sense that they tell you a number someone actually experienced and they put an upper bound on that experience. they are bad because they won’t tell you is how the
208.
by dastbe 5y ago
it has a pipeline that dumps people into positions that they may or may not be able to do. employers that promote based on work can and are invested in developing talent but ensure that people can actually do the role they are hired into.
209.
by dastbe 5y ago
> There’s always a balance. “If you’re at an organization who demands you do the next level job before promotion, they’re cheating you of the pay difference for a period of time.” this is true, though i would say there are two things her
210.
by dastbe 5y ago
if youre in an organization that promotes based on potential to do the work of the next level rather than demonstrating you can do the work of the next level, you should run

More ›