alexandrnikitin

4 ms·

1.
by alexandrnikitin 4y ago
I wrote a post on how to measure perf impact and pitfalls of THP some time ago. I hope it helps https://alexandrnikitin.github.io/blog/transparent-hugepages...
2.
by alexandrnikitin 9y ago
> Bad advise... That is exactly the reason I wrote the post! Those advice are based on specific use case, bug or outdated kernel. The jemalloc (Digital Ocean post) case is a good example, it just doesn't (didn't) know about THP
3.
by alexandrnikitin 9y ago
> My github portfolio is a means for me to determine if a potential employer is suitable for me, not the other way around. It's not a test for me to pass, it's a test for a potential employer to pass. Nice! Agree with that.
4.
Transparent Hugepages: measuring the performance impact(alexandrnikitin.github.io)
2 points by alexandrnikitin 9y ago|0 comments
5.
by alexandrnikitin 9y ago
> But this guy is writing server software. Micro-optimizing on the server side the way he is doing is silly. Khm... What if you have millions of requests per second with tight latency requirements measured in milliseconds; and a bunch of
6.
by alexandrnikitin 9y ago
It depends. Usually do nothing if that traffic is very low. There's no reliable way to do that. Honeypots and behavior analysis are very useful here.
7.
by alexandrnikitin 9y ago
I agree with you, usually you shouldn't optimize to the point when code quality starts suffer. It's all about trade offs. If you have one or two hundred servers and millions of RPS then it could be reasonable. Or coming back to th
8.
by alexandrnikitin 9y ago
I intend to write a separate blog post about low-overhead production monitoring (not sure when it happen though)
9.
by alexandrnikitin 9y ago
Awesome! The talk is great! It would be really interesting to try it. Thanks for sharing.
10.
by alexandrnikitin 9y ago
I'm afraid I can't do that because of proprietary data. I think I can come up with analogous tests using open data. I'll let you know ;)
11.
by alexandrnikitin 9y ago
I have that feeling too. But I found it harder to implement, especially with fallback failure references.
12.
by alexandrnikitin 9y ago
Yes, using honeypots is one of the ways to identify bots. But that wasn't the focus of the post. I'll add some clarification.
13.
by alexandrnikitin 9y ago
I doubt that exactly that will work. There are tens of thousands of different UAs (maybe 100K). Perhaps some kind of tiny (few CPU cache lines) cache for most popular UAs could help. But again: measure, measure, measure :)
14.
by alexandrnikitin 9y ago
Yes, you're right. There are many ways to block robots: IP, UA, behaviour analysis. An advertising company has to have UA based filtering to be compliant with standards. However, the focus of the blog post is on performance rather on h
15.
High-performance .NET by example: Filtering bot traffic(alexandrnikitin.github.io)
228 points by alexandrnikitin 9y ago|50 comments
16.
Bloom filter for Scala, the fastest for JVM(alexandrnikitin.github.io)
6 points by alexandrnikitin 10y ago|0 comments
17.
Running Java inside a Windows container on a Windows server(alexandrnikitin.github.io)
2 points by alexandrnikitin 10y ago|0 comments
18.
Hoisting in .NET Examples(alexandrnikitin.github.io)
1 points by alexandrnikitin 11y ago|0 comments