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

Scylla 3.0 announced – Secondary Indexes, Cassandra 3.0 compatibility and more

scylladb.com

Post reply on HN
Print view View original

11–18 of 18 posts

Re: Scylla 3.0 announced – Secondary Indexes, Cassandra 3.0 compatibility and more

#11

Post by atombender »

The announcement is a little misleading — they're announcing two different things, not one:

* Scylla 3.0, which adds secondary indexes and materialized views; and

* OLTP and OLAP features, which are not ready.

I'm struggling to find any clear information on what's implied by OLTP, but from the roadmap [1], it looks like they're just adding Cassandra's LWTs, not ACID transactions. Last I heard, you couldn't build ACID on top of Cassandra/Scylla, since row updates across multiple keys cannot be done atomically. Calling this OLTP seems a little misleading.

[1] https://www.scylladb.com/product/technology/scylla-roadmap/

Re: Scylla 3.0 announced – Secondary Indexes, Cassandra 3.0 compatibility and more

#12

Post by mcms »

atombender wrote: post #11

The announcement is a little misleading — they're announcing two different things, not one: * Scylla 3.0, which adds secondary indexes and materialized views; and * OLTP and OLAP features, which are not ready. I'm struggling to find any clear information on what's implied by OLTP, but from the roadmap [1], it looks like they're just adding Cassandra's LWTs, not ACID transactions. Last I heard, you couldn't build ACID…

Actually, Scylla 3.0 is still RC [1] and as mentioned in the article will be released later this month.

And I don't think OLTP/OLAP is about LWTs. I think they are just talking about Scylla's ability to process normal updates in a timely manner while doing OLAP queries.

And you're right about ACID transactions. Multiple rows across two or more partitions can not be updated atomically, even with LWTs, but it is possible for a single partition.

[1] https://github.com/scylladb/scylla/releases

Re: Scylla 3.0 announced – Secondary Indexes, Cassandra 3.0 compatibility and more

#13

Post by erikrothoff »

nemothekid wrote: post #10
erikrothoff wrote: post #4

I wad looking into Scylla yesterday so this is really cool timing! Does anyone have any experience with just switching out (Apache) Cassandra with Scylla?

I've managed medium sized Cassandra clusters (~18 nodes) for a couple years (since v1.2). At a new job, last year, I replaced the Cassandra cluster with Scylla. Performance was great, as it says on the in, but as an operator: - No GC Tuning. It just works out of the box and I don't sleep over requests being dropped because of a multi-second GC. - Compactions are fast. Compactions in Cassandra generate a lot of garbag…

When you did the switch did you just swap out the software on each node, or did you provision new nodes with Scylla, adding them to the cluster, and removing old nodes with Cassandra?

Re: Scylla 3.0 announced – Secondary Indexes, Cassandra 3.0 compatibility and more

#14

Post by nemothekid »

erikrothoff wrote: post #13
nemothekid wrote: post #10

Earlier quoted context omitted.

I've managed medium sized Cassandra clusters (~18 nodes) for a couple years (since v1.2). At a new job, last year, I replaced the Cassandra cluster with Scylla. Performance was great, as it says on the in, but as an operator: - No GC Tuning. It just works out of the box and I don't sleep over requests being dropped because of a multi-second GC. - Compactions are fast. Compactions in Cassandra generate a lot of garbag…

When you did the switch did you just swap out the software on each node, or did you provision new nodes with Scylla, adding them to the cluster, and removing old nodes with Cassandra?

I don't believe you can have a mixed Scylla/Cassandra cluster. In our case we cloned our drives and started a whole new cluster with Scylla. When we were happy with Scylla and done the appropriate testing, we shutdown the Cassandra cluster.

Re: Scylla 3.0 announced – Secondary Indexes, Cassandra 3.0 compatibility and more

#15

Post by atombender »

mcms wrote: post #12
atombender wrote: post #11

The announcement is a little misleading — they're announcing two different things, not one: * Scylla 3.0, which adds secondary indexes and materialized views; and * OLTP and OLAP features, which are not ready. I'm struggling to find any clear information on what's implied by OLTP, but from the roadmap [1], it looks like they're just adding Cassandra's LWTs, not ACID transactions. Last I heard, you couldn't build ACID…

Actually, Scylla 3.0 is still RC [1] and as mentioned in the article will be released later this month. And I don't think OLTP/OLAP is about LWTs. I think they are just talking about Scylla's ability to process normal updates in a timely manner while doing OLAP queries. And you're right about ACID transactions. Multiple rows across two or more partitions can not be updated atomically, even with LWTs, but it is possib…

But OLTP is online transactional processing. Even with LWT, you arguably just can't do OLTP with Cassandra/Scylla, not without transactions.

Re: Scylla 3.0 announced – Secondary Indexes, Cassandra 3.0 compatibility and more

#16

Post by thdxr »

atombender wrote: post #15
mcms wrote: post #12

Earlier quoted context omitted.

Actually, Scylla 3.0 is still RC [1] and as mentioned in the article will be released later this month. And I don't think OLTP/OLAP is about LWTs. I think they are just talking about Scylla's ability to process normal updates in a timely manner while doing OLAP queries. And you're right about ACID transactions. Multiple rows across two or more partitions can not be updated atomically, even with LWTs, but it is possib…

But OLTP is online transactional processing. Even with LWT, you arguably just can't do OLTP with Cassandra/Scylla, not without transactions.

OLTP traditionally refers to the type of updates you do during normal use of your application. Frequent, small and incremental updates vs large sweeping reads and aggregations (OLAP)

The T does stand for transactional but in this context it is not referring to transactions in the database sense

Re: Scylla 3.0 announced – Secondary Indexes, Cassandra 3.0 compatibility and more

#17

Post by duarten »

nemothekid wrote: post #14
erikrothoff wrote: post #13

Earlier quoted context omitted.

When you did the switch did you just swap out the software on each node, or did you provision new nodes with Scylla, adding them to the cluster, and removing old nodes with Cassandra?

I don't believe you can have a mixed Scylla/Cassandra cluster. In our case we cloned our drives and started a whole new cluster with Scylla. When we were happy with Scylla and done the appropriate testing, we shutdown the Cassandra cluster.

[Scylla employee] Indeed a mixed cluster isn’t supported, since the inter-node protocols are incompatible.

Re: Scylla 3.0 announced – Secondary Indexes, Cassandra 3.0 compatibility and more

#18

Post by duarten »

mcms wrote: post #12
atombender wrote: post #11

The announcement is a little misleading — they're announcing two different things, not one: * Scylla 3.0, which adds secondary indexes and materialized views; and * OLTP and OLAP features, which are not ready. I'm struggling to find any clear information on what's implied by OLTP, but from the roadmap [1], it looks like they're just adding Cassandra's LWTs, not ACID transactions. Last I heard, you couldn't build ACID…

Actually, Scylla 3.0 is still RC [1] and as mentioned in the article will be released later this month. And I don't think OLTP/OLAP is about LWTs. I think they are just talking about Scylla's ability to process normal updates in a timely manner while doing OLAP queries. And you're right about ACID transactions. Multiple rows across two or more partitions can not be updated atomically, even with LWTs, but it is possib…

[Scylla employee] We’re bootstrapping our LWT implementation now and we do want to take it much further, with multi-partition transactions.
Post reply on HN