halbritt

23 ms·

241.
by halbritt 8y ago
Self-hosting doesn't scale? Have you done the math? If so, I'd be curious to see it. There are a number of reasons that self-hosting doesn't make sense which have very little to do with scale and more to do with the lack of
242.
by halbritt 8y ago
There's a world of organizations between Netflix and Twitter scale and small startups with 2-3 person infrastructure teams in which Kubernetes makes good sense. At some point, it becomes cost effective to have a two-pizza team of engin
243.
by halbritt 8y ago
I'm generally a proponent of DIY, this doesn't make sense. When an org needs to scale, it makes sense to cultivate skill in the underlying infrastructure. On day 1, serverless makes a lot of sense because it encourages developme
244.
by halbritt 8y ago
Obviously, there's a huge cost associated with the learning curve, but this the part of the reason that Kubernetes is so attractive. It abstracts away the underlying infrastructure, which is nice. At any kind of scale, though, one is
245.
by halbritt 8y ago
In the immortal words of @vgill, "You own your availability."
246.
by halbritt 8y ago
That's terrible best-practice advice. My best-practice advice is to do the math. What is the margin on infrastructure vs. the acquisition and management costs of the engineers necessary to operate the infrastructure. Serverless doesn&
247.
by halbritt 8y ago
I'm not defending the practice, but it's disingenuous to suggest that it solves nothing. Ostensibly it's a means by which the city can determine which programs are most effective at getting the homeless population off the st
248.
by halbritt 8y ago
If it's a given that cloud providers are aiming to get lock-in and increase margins, then it makes sense to do the necessary math to avoid that.
249.
by halbritt 8y ago
I really appreciate this part: Some of these firms realize they need to renovate their IT but they attempt to do this through hiring “Digital Transformation” roles which do little but frustrate the revolving door of hires becau
250.
by halbritt 8y ago
They call this "lift and shift". It's a terrible idea, but cloud providers love it because they end up getting paid a lot of money for idle capacity.
251.
by halbritt 8y ago
I don't think it's that hard to reach a scale where DIY makes sense. As was stated up above, the AWS kafka offering isn't complete. One could do something similar with a cloud formation template. The ElasticSearch offering
252.
by halbritt 8y ago
I've given it a test drive and highly approve. It is very polished. Kinda spendy, though.
253.
by halbritt 8y ago
I'm using zstd with my kafka implementation and it works great. v2.1 is required. We're using librdkafka, which works, but I believe we had to compile it in to get that and idempotent producer.
254.
by halbritt 8y ago
My question, perhaps poorly articulated was why would someone focus on improving single-node performance vs. cluster-wide performance? I wasn't implying that one would want to run a single-node cluster.
255.
by halbritt 8y ago
There's a JIRA issue that's been open for three years for kafka to support openssl with JNI: https://issues.apache.org/jira/browse/KAFKA-2561 Not sure it's going to get done at this point. I'v
256.
by halbritt 8y ago
One wonders why this was downvoted.
257.
by halbritt 8y ago
For some use cases, it absolutely does. I can't remember the exact use case, but a friend of mine is adopting kafka with an end-to-end latency goal of 100ms. Absolutely requires SSDs and for the producers and consumers to be colocate
258.
by halbritt 8y ago
The problem is that it's not implemented in the native library. That's why the throughput is so abysmal.
259.
by halbritt 8y ago
I understand that you can increase single-broker throughput, but one wonders what use case requires that? That'd be warranted in the case of single-partition topics which are necessary to preserve strict ordering. If one has ability
260.
by halbritt 8y ago
Sure, sure, but did you get that kind of throughput with kafka on java? Most folks (Confluent, for example) are still on java 8 which has some very severe limitations with regards to throughput.
261.
by halbritt 8y ago
With TLS?
262.
by halbritt 8y ago
The AWS kafka product is a non-product. One would have to reproduce the same effort there as anywhere else. 60MB/s per broker Took me about two hours worth of testing to reach that same conclusion. I like the Dropbox approach thoug
263.
by halbritt 8y ago
It's not so much that anybody has been convinced, but that Apple themselves are incentivized to preserve the privacy of their customers to some degree.
264.
by halbritt 8y ago
Speaking of folks operating with no safety devices other than that which is between his ears. I love that channel.
265.
by halbritt 8y ago
I've worked with folks like that. They usually do more harm than good and generally need to be managed out or into roles more suitable for how they like to work.
266.
by halbritt 8y ago
Precisely. A decent new watch isn't going to have a defect. The parent referred to a theoretical $500 watch. Let's take something like a Hamilton Khaki Field Automatic. Decent watch, very handsome, and usually comes in around $
267.
by halbritt 8y ago
It's probably worth clarifying why it's obvious that he was an adherent of George Daniels. In his book, and in every interview I've seen of him, he speaks at length about the value of a watch maker being able to make every c
268.
by halbritt 8y ago
This guy is obviously an adherent of George Daniels. His story is fascinating. He produced likely one of the only significant inventions in mechanical watches in the last century, the coaxial escapement. The book shown in the OP is wonde
269.
by halbritt 8y ago
Warranties cover defects which are few. A $500+ watch will need servicing after a few years, which is not covered and will generally run a few hundred dollars. Not really a "big risk".
270.
by halbritt 8y ago
The article disagrees with your assertion: Some 1.2bn people have been lifted out of extreme poverty since 1990

More ›