Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Clustrix, a distributed SQL DB, launches on AWS

clustrix.com

Post reply on HN
Print view View original

11–20 of 42 posts

Re: Clustrix, a distributed SQL DB, launches on AWS

#11

Post by sergei »

indeyets wrote: post #7
polskibus wrote: post #5

does anyone know more about their "unique parallel query approach that generates distributed query plans and query fragments for speed and scale." ? I'm not going to trust a database which internals (or at least core mechanisms) I don't understand. Any links to papers perhaps?

I wonder if that's just mongodb-style sharding

It's different from a typical sharding approach (including what MongoDB does). In their model, you take a single key and distribute your data using that key (e.g. user_id). The problem surfaces when you look at secondary indexes.

If you have a secondary index say on user_location, and you want to query by that index, you don't know which shard to go to. So you end up broadcasting.

Another problem is enforcing unique index constraints.

With Clustrix, every table and index gets its own distribution.

So if you have a schema like this:

foo(a, b, c, d) unique idx1(b,c) idx2(d)

Clustrix treats each table and index as a different distribution. So if I need to look something up by d, I know exactly which node has the data. I can also enforce index uniqueness.

Re: Clustrix, a distributed SQL DB, launches on AWS

#12

Post by tedchs »

indeyets wrote: post #3

So, there's Amazon Redshift (which is cloud-postgres) https://aws.amazon.com/redshift/ and now there's Clustrix (which is cloud-mysql)

In case you were saying Redshift is literally PostgreSQL, it appears it's actually ParAccel:

http://www.informationweek.com/software/information-manageme...

Re: Clustrix, a distributed SQL DB, launches on AWS

#15

Post by blasdel »

indeyets wrote: post #3

So, there's Amazon Redshift (which is cloud-postgres) https://aws.amazon.com/redshift/ and now there's Clustrix (which is cloud-mysql)

Redshift is a column-store database for big analytics workloads and has a lot of very big established competitors, though the cloud space for that is mostly startups. There's a number of products that are really storage backends for postgres, or use its SQL parsing frontend.

What's different about Clustrix compared to all the other distributed databases branding themselves as "NewSQL" is that it's intended for totally traditional OLTP workloads (though it can distribute OLAP queries across the cluster too as a bonus).

Disclaimer: I work there

Re: Clustrix, a distributed SQL DB, launches on AWS

#16

Post by polskibus »

sergei wrote: post #11
indeyets wrote: post #7

Earlier quoted context omitted.

I wonder if that's just mongodb-style sharding

It's different from a typical sharding approach (including what MongoDB does). In their model, you take a single key and distribute your data using that key (e.g. user_id). The problem surfaces when you look at secondary indexes. If you have a secondary index say on user_location, and you want to query by that index, you don't know which shard to go to. So you end up broadcasting. Another problem is enforcing unique…

I saw claims that Clusterix is very good for OLAP applications. Can you shed more light on it? Does it support materialized views for instance (to speed up calculating aggregation on higher levels) ?

Re: Clustrix, a distributed SQL DB, launches on AWS

#17

Post by dpollack »

snowmaker wrote: post #6

Does anyone here have direct customer experience with Clustrix? If so, I would be very interested in hearing honest customer feedback.

We have been using clustrix at AOL for quite a while. You can check their site for our testimonial. We find it to be reliable and easy to use. It requires so little care and feeding we haven't had a full time DBA on our application for several months. We've done consolidations from a sharded MySQL environment to clustrix using nothing more complex than MySQL replication. We also been able to make online schema changes without locking the DB on very large tables in excess of 400 million rows. Support is quite proactive as well both for regular release upgrades as well as features additions and minor bug fixes. We've been in production for more than 2 years and other than minor issues Clustrix has been trouble free.

Re: Clustrix, a distributed SQL DB, launches on AWS

#19

Post by mjw00 »

polskibus wrote: post #16
sergei wrote: post #11

Earlier quoted context omitted.

It's different from a typical sharding approach (including what MongoDB does). In their model, you take a single key and distribute your data using that key (e.g. user_id). The problem surfaces when you look at secondary indexes. If you have a secondary index say on user_location, and you want to query by that index, you don't know which shard to go to. So you end up broadcasting. Another problem is enforcing unique…

I saw claims that Clusterix is very good for OLAP applications. Can you shed more light on it? Does it support materialized views for instance (to speed up calculating aggregation on higher levels) ?

We don't support materialized views at the moment, though I can't think of a reason why we couldn't support them if needed. A big strength for our OLAP performance is simply having CPU and memory resources that scale with storage and a query planner that is smart enough to take advantage of those resources.
Post reply on HN