[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[DISCUSS] Cassandra and future Java

I'd like to bring up the C*/Java discussion again. It's been a while since we've discussed this.

To me it sounds like there's still the question about which version(s) of Java we want to support beginning with C* 4.0.

I assume, that it's legit (and probably very necessary) to assume that OpenJDK is now (i.e. after Java 6) considered as "production ready" for C*. The public (and legal and free) availability of Oracle's Java 8 will end in January 2019 (unless you're using it privately on your desktop). Java 9 and 10 are not a thing, as both will be EOL when the C* 4.0 branch is about to be cut. The most recent available Java version will be 11, which is meant to be publicly available from Oracle until March 2019 and should get LTS support for OpenJDK 11 from major Linux distros (RHEL and derivates, Ubuntu, Azul Zulu).

(Side note: adoptopenjdk is different here, because it does not include the patch version in the version banner (java.version=1.8.0-adoptopenjdk), so difficult to check the minimum patch version on startup of C*.)

(Attn, rant: I'm not particularly happy with the new release and support model for Java, because developing something now, that's about to release end of the year on a Java version that has not even reached feature-complete status, is, gently speaking, difficult. But sticking to an "antique" Java version (8) has its own risks as well.)

I'm silently ignoring any Java release, that's not aimed to get any LTS(-ish?) support from anybody - so only Java 8 + 11 remain.
There are generally three (IMO legit) options here: only support Java 8,
only support Java 11, support both Java 8 and Java 11. All three options
have a bunch of pros and cons.

Option 1, only Java 8: Probably the safest option. Considering the potential lifetimes of Java 8 and C* 4.0, even the most enthusiastic maintainers may stop backporting security or bug fixes to OpenJDK 8. It might not be an issue in practice, but if there's for example a severe issue in the SSL/TLS area and nobody fixes it in 8, well, good luck.

Option 2, only Java 11: The option with the most risks IMO. Java 11 is not even feature complete, and there a bunch of big projects that still may make it into 11 (think: Valhalla). There's no guarantee whether the C* code or any included library will actually work with Java 11 (think: if it works now, it may not work with the final Java version). However, it leaves the door wide open for all the neat and geeky things in Java 11.

Option 3: both 8 + 11: The idea here is to default to Java 8, but the code also runs on 11. It leaves the option to benefit from optimizations that are only available on 11 while maintaining the known stability of 8. Initially, only the combination of C* 4.0 + Java 8 would be labeled as "stable" and the combination of C* 4.0 + Java 11 as "experimental". But it gives us time to "evaluate" 4.0 on 11. When we have enough experience with 11, C* on 11 can be labeled as "stable" as well. The downside of this "hybrid" is, that it's a bit more difficult to introduce features, that depend on 11.

I think, 3) gives the best of both worlds: stability of 8 and an upgrade path to 11 in the future, that people can actually test with C* 4.0. Happy to make the patch for #9608 ready for option 3. But it would be great to get a consensus here for either option before we review #9608 and commit it.

Another proposal, for both options 1+3: Raise the minimum supported version of 8 for C* 4.0 to something more recent than 8u40, which is quite from the stone-age. It could be 8u171 or whatever will be recent in autumn.


Robert Stupp

To unsubscribe, e-mail: dev-unsubscribe@xxxxxxxxxxxxxxxxxxxx
For additional commands, e-mail: dev-help@xxxxxxxxxxxxxxxxxxxx