arynda

5 ms·

1.
by arynda 3y ago
If you want an actual startup-oriented bank that has their shit together (e.g. they employ sweep accounts and by default you have $5M of FDIC insurance), mercury.com has been excellent. Great tech, great UI/UX, great app, transfers are
2.
by arynda 3y ago
I miss Topcoder. Funny story I went back through all my old SRMs and found one where I was placed in the same room with none other than mzuckerberg. He even unsuccessfully challenged my code and I in turn challenged his with a test he hadn’
3.
by arynda 4y ago
any chance you have a link to a good readme on setting up single-node docker swarm?
4.
by arynda 4y ago
> ClickHouse performs what is known as PREWHERE processing > p.s. Scans are ridiculously fast. Good point, I should have mentioned this was basically a worst-case scenario for Clickhouse as the data layout is totally random (same appr
5.
by arynda 4y ago
Not in particular sorry, most of the good content I've found is on Altinity [1] and Alibaba's technical blogs [2][3]. These tend to be mostly focused on how the data itself is stored and how to use Clickhouse, but don't reall
6.
by arynda 4y ago
Comparison on Clickhouse, also runs in about 30-40ms, however there's no indexing being used and this is a full-table scan. create table if not exists test_table ( id UInt64, text1 String, text2 String
7.
by arynda 4y ago
The same way a spoken language is a communication tool and doesn't restrict what is discussed.
8.
by arynda 4y ago
https://asciiflow.com/
9.
by arynda 4y ago
That's cool. I've been thinking through a similar design. Would this be an accurate description of your approach? queues hold jobs, the scheduler/dispatcher handles load-balancing and rate-limiting, pushing jobs down to worke
10.
by arynda 4y ago
How did you end up solving the rate limiting? Were the workers themselves responsible for coordination via a semaphore (e.g. there's already 25 of us, don't consume from this queue) or did you solve it on the scheduling side and o