HNHacker News
TopNewBestAskShowJobs

jjirsa

793 karma · joined December 10, 2013

Databases, storage, infrastructure
submissionscomments
jjirsa··on Benchmark results: Cassandra 4.0, 3.11, Scylla 4.4 [video]
Vectorized/RedPanda for Kafka compatible service, at least.

RedPanda seems to be a better starting point to be honest - a lot of the market share for Cassandra doesn't actually need the extra tps of Scylla. Some customers do, but many many many do not.

jjirsa··on Ask HN: What could a modern database do that PostgreSQL and MySQL can't
Sharding is not the same as natural clustering, because eventually you’ll need to reshard and then you’ll be writing a lot more code.
jjirsa··on Bitcoin hits 20K
A currency that is used for real, genuine transactions and also happens to burn energy to keep the world economy honest? Can argue that's worthwhile.

A pyramid scheme that has electricity consumption of bitcoin? Strictly a net loss for the world.

jjirsa··on Why should I do production support?
> You aren't looking for the hardest problems, you're looking for the problems your users hit the most that an engineer could reduce in the product.

This is the hard part of software.

jjirsa··on Why should I do production support?
3 is a common mistake, like the one engineers make when they pretend that the hard part of programming is typing
jjirsa··on Why should I do production support?
Why should a company hire two lines of support to shield an engineer from bugs they introduce?
jjirsa··on Graviton Database: ZFS for key-value stores
Define “at scale”
jjirsa··on Graviton Database: ZFS for key-value stores
> You need a team probably

The cardinal rule of database development:

http://www.dbms2.com/2013/03/18/dbms-development-marklogic-h...

jjirsa··on TikTok’s potential $588 per user valuation in Microsoft deal
Who created this lie that Microsoft struggled with consumer?

Xbox was always consumer and is massive. A huge chunk of the Windows line was always about consumers. The office line straddled both. Hotmail was consumer first.

jjirsa··on Launch HN: Datafold (YC S20) – Diff Tool for SQL Databases
For Cassandra: https://github.com/apache/cassandra-diff

(Designed for correctness testing at petabyte scale)

jjirsa··on What to know to get promoted quickly in Big Tech
Not as silly as you think
jjirsa··on Apple is looking for engineers to convert its code from C to Rust
> Following a very successful first foray into Rust we are migrating an established codebase from C to Rust, and building new functionality primarily in Rust.

That doesn't say the job is ONLY converting code. It says that code was converted, and new code will be in Rust.

jjirsa··on Guide to running Elasticsearch in production
It’s highly concurrent but not pauseless
jjirsa··on State of Postgres
If 9% contributed code, that sounds like a really biased sampling
jjirsa··on RDS Pricing Has More Than Doubled
Clearly RDS prints money - doesn’t take a genius to realize AWS has been trying to duplicate that success with every other OSS db-aaS they’ve launched since.

Given its success, lack of pricing pressure makes sense - until users go away, there’s little reason to decrease margins.

jjirsa··on Amazon Launches Managed Cassandra Service
Cassandra is an Apache project and will always have an Apache license.

It’s not a startup-driven product that is worried about AWS as a competitor.

jjirsa··on Amazon Launches Managed Cassandra Service
I don't work for AWS, but I do run Cassandra at scale for a living, and this line:

> Considering the kind of resources Apache Cassandra requires to run

Makes no sense. There is plenty of talent at AWS to run Cassandra proper on AWS for less money than they're currently charging for Dynamo. This was true before this announcement - it's how people like Instaclustr stay in business (containerized cassandra as a service on EC2).

jjirsa··on Amazon Launches Managed Cassandra Service
> Cassandra was hard to manage and after many years of our team using it still had random spikes

A lot of that is getting better over time. A non-trivial percentage of contributions are coming from companies that are only looking to improve their operational story, not adding new features.

Most of the talks at NGCC (Next generation cassandra conference) focused on operational improvements - it's something we care a lot about (myself especially).

jjirsa··on Amazon Launches Managed Cassandra Service
Biased as a cassandra committer, but:

- Feature sets aren't the same yet

- History / maturity

- License

- Actual savings dont tend to match proposed savings

- Development isn't driven by a startup that AWS can kill with an announcement like this.

jjirsa··on Amazon Launches Managed Cassandra Service
AGPL and immature.
jjirsa··on How to run your own mail server (2017)
GMail being at fault isn't going to make up for a missed invitation (or RSVP) to a special event, or not getting a note of congratulations on an accomplishment, or not getting a message of condolences.

Assigning blame is done after the fact to make up for a mistake.

Just avoid the mistake.

jjirsa··on How to run your own mail server (2017)
> I don't care if people don't get my mail

This is the type of nonsense I expect from HN. Well done.

jjirsa··on How to run your own mail server (2017)
It's not that they dont know or don't care - it's that not being super aggressive with filtering caused people to almost abandon email a decade ago because spam became unbearable for most people.

Those "specialized tools" make email viable. Without it, everyone would be spoofing everything and you wouldn't use email anymore.

jjirsa··on How to run your own mail server (2017)
EVERYONE did this for decades.

Then EVERYONE stopped because it's awful. Modern ISPs don't like SMTP floating around, (DKIM, DMARC, SPF) are a pain to maintain over time, exploits will happen, and your free off-the-shelf virus and spam filters suck compared to the big companies seeing billions of messages per day.

jjirsa··on WeWork All Hands Transcript
Dev finance group and comment at a precious all hands, I was guessing person called out previous CEO in public.
jjirsa··on WeWork All Hands Transcript

    Employee 3

    Hi, my name is [redacted]. (audience applause)

    Marcelo Claure

    Well first you’ve got to tell me why the people like you so much? (audience laughs)

    Employee 3

    Because I asked a question at the last All-Hands meeting.

    Marcelo Claure

    Okay.

    Employee 3

    I’m with the Development Finance Group, I do data analytics and reporting and today’s my one-month anniversary at WeWork. (applause)

    Marcelo Claure

    You’re very popular for one month.

Not asking anyone to dox employee 3, but I’d really love to know what the backstory there is.
jjirsa··on Flash Is Responsible for the Internet's Most Creative Era
All of that and no 2Advanced
jjirsa··on Can someone explain Merkle Trees to me in a practical sense?
An easy example of this is the implementation in Apache Cassandra

In Cassandra, we use MTs during active anti-entropy repair, where we have N replicas of some data and want to make sure they’re in sync

The repair command will repair a range of data - in real terms, it’s repairing all keys between two tokens, where the token is the murmur3 hash of the key. The whole cluster covers 2^64, but a typical repair will repair perhaps 2^40 or smaller.

When you’re repairing 2^40 keys, you don’t want to compare each of them one by one - if they match, you’re sending a ton of data across the network for no reason. Instead, if each replica builds a needle tree representing the data for that range, you can not only tell quickly if they’re in sync (comparing the roots), but you can descend to identify the minimal set of data to stream between replicas.

The Cassandra implementation isn’t a textbook implementation of the concept, but it’s easy to think about.

jjirsa··on Ask HN: Resources for Learning about Databases
Alex’s book: https://www.databass.dev/
jjirsa··on DynamoDB Compatibility Layer for Apache Cassandra
Git history suggests it’s been around for a year. It’s not super interesting really, clearly there’s a poke here that someone felt this feature was worth a press release and their more established competitor just “meh, here”’d them.

Also, cql is now implemented by like 5-6 nosql databases and is becoming the implicit standard (Cassandra, cosmosdb, datastax, yugabyte, etc).

← PreviousPage 2 of 8Next →