The latest expert opinions, articles, and guides for the Java professional.
It’s an exciting time for the LiveRebel team–we released LiveRebel 3.1 a few days ago and so far nothing has broken! Kidding aside, releasing software is a difficult task, and that why we built LiveRebel: so that any size software shop could afford out-of-the-box continuous delivery with multi-platform deploys, release orchestration, zero downtime rollouts and automated rollbacks.
LiveRebel 3.1 focuses on performance enhancements (i.e lots of servers) and security improvements, which we think makes LiveRebel a bit more mature. Plus, our new rolling update strategy is now available for all servers, while we’ve completely moved away from hotpatching. Existing customers can easily upgrade using this migration guide to make the transition from previous versions as smooth as possible.
So if you’d like to learn more about this stuff, we’re having a live webinar on Aug 7th.
Over the past three months we’ve had to ask AWS to raise our instance quota three times, up to 1000 nodes. AWS sales reps are definitely happy, but more importantly, all our customers should be too. We’ve eliminated countless of bottlenecks and it shows–LiveRebel feels much more responsive across the board and servers connect in a snap.
In most setups, LiveRebel is hosted in the internal network, while more and more servers live externally on PaaS or cloud environments. With this in mind, user interface and servers now talk on separate ports. System administrators can enforce different policies for the mostly internal and external communication.
We’ve also made improvements for setting up LiveRebel itself. You can configure the protocol–http or https–and use your own SSL certificates. On top of that we’ve made sure that LiveRebel works behind a front-end proxy through X-forwarded headers.
Now roll out apps without session drain(!)
As we announced at the start of the year, LiveRebel 3.1 has completely moved away from hotpatching. Rolling updates with session drain have given the same zero downtime benefits without any “magic” that doesn’t really belong in a production environment.
But waiting for sessions to drain still requires delays, so as a make up gift, we introduced a new strategy of rolling out updates without session drain. LiveRebel releases apps by some given number of nodes at a time, while the front-end proxy can route requests to available nodes. It also works for applications where http traffic doesn’t exist–like in standalone Java apps and file apps.
How to try LiveRebel free
Check out LiveRebel 3.1 and let us know what you think–you can learn a lot at our upcoming webinar, so until then we’re eager to hear your feedback! Tweet us at @liverebel or email email@example.com with any questions or comments.
No comments yet.
Leave a comment