jclouds community is an international group with over 100 contributors since we started in early 2009. The idea of holidays vary, but we decided Mistletoe as an appropriate codename for jclouds 1.5.4. Here's why.

Chef and Karaf

We've noticed an attraction between systems that use chef and those that use karaf. For example, it isn't uncommon to have a deployment pipeline that incorporates the two. Since jclouds has subproject for both chef and karaf, it seemed natural to cozy them together. Thanks to Ioannis and Ignasi, karaf and chef are now a couple!

For example, the following syntax can be used in karaf directly, or via the jclouds cli to start a node in any supported cloud and bootstrap chef recipes.

karaf@root>node-create --adminAccess --recipe chef/java::openjdk karaf

What else is in 1.5.4?

We've a number of other highlights in jclouds 1.5.4. Many thanks to the code contributors jclouds, jclouds-chef, jclouds-karaf, and jclouds-cli for putting them together. Also to our review team for helping ensure quality: Andrew Gaul and Matt for code, and Becca on docs.

Rackspace Cloud Load Balancers

Everett modernized support for rackspace-cloudloadbalancers-us and uk is now available. This includes syntactic sugar such as auto-pagination and easy transforms.

FluentIterable<HostAndPort> sockets = clb.getLoadBalancerApiForZone(zone).list().concat().transform(converter);

Nova Server Diagnostics

Some implementations of Nova have diagnostic information available. Thanks to Leander, you can now discover this capability at runtime.

Optional<Map> diagnosticInfo = novaApi.getServerApiForZone("az-1.region-a.geo-1").getDiagnostics(serverId);

S3 Multi-Object delete

Maginatics have BlobStore containers with over a billion objects in them. S3 containers (buckets) can now be cleared with 1/1000 the requests, thanks to Andrei's additional support of S3 multi-delete.

DeleteResult result = s3Api.deleteObjects(container, keys);
assertEquals(result.size(), keys.size());

What's next

We are working on 1.6, an effort more efficient and resilient. One tool we are evaluating for this is Netflix Hystrix. We are also working on 1.5.x releases, which are likely to include google compute engine and virtualization support. If interested in helping us out, hop onto IRC freenode #jclouds or join our dev group.


Comments powered by Disqus