DataStax Support Forums » Tag: cluster

DataStax Support Forums » Tag: cluster - Recent Posts http://www.datastax.com/support-forums/tags/cluster Software, Support, and Training for Apache Cassandra en-US Tue, 21 May 2013 21:41:08 +0000 http://bbpress.org/?v=1.0.3 <![CDATA[Search]]> q http://www.datastax.com/support-forums/search.php joaquin on "aws ami" http://www.datastax.com/support-forums/topic/aws-ami#post-2959 Thu, 12 Jul 2012 06:31:03 +0000 joaquin 2959@http://www.datastax.com/support-forums/ Hello pnegri, You would ideally want to launch a new cluster with the same number of nodes that you are planning on adding. Yes, you would want to merely modify all the tokens, seed ips, and set auto_bootstrap to true. Stop the service, clear the data at /raid0/cassandra/*, then start each node with 2 minutes in between, along with anything else I may be missing from the docs. I used to have this option, but we deprecated it due to an expanding QA rubric. However, the typical use case for the AMI is something of a more easy, friendly, well-configured developer tool. Once you get to expanding a cluster, that's more of an "advanced" scope than what the AMI was meant to handle. The expanding is also simple enough to where most cases are fairly straight forward. However, the edge cases where you're adding a new datacenter would be a bit tougher to code without some sort of manual intervention either way. Reflector.php's description can be found here: <a href="https://github.com/riptano/ComboAMI/blob/2.2/FILES.md#reflectorphp" rel="nofollow">https://github.com/riptano/ComboAMI/blob/2.2/FILES.md#reflectorphp</a> Tokentool is a script that provides the computation for the tokens when the cluster boots up. Cheers, Joaquin pnegri on "aws ami" http://www.datastax.com/support-forums/topic/aws-ami#post-2936 Wed, 11 Jul 2012 17:36:44 +0000 pnegri 2936@http://www.datastax.com/support-forums/ Hey Joaquin, what do you mean by "In the AWS Management Console, create a cluster with the number of nodes you want add, as described above."? Create a new cluster with a different name from the running one? Stop all, and then reconfigure and join? Wouldnt be better if we could supply our tokens when launching the new AMIs? Something like a manager or a URL to auto assign these? Also, whats the purpose of reflector.php and tokentook? Thank you dan on "3 Node Cluster RF=3" http://www.datastax.com/support-forums/topic/3-node-cluster-rf3#post-2823 Mon, 09 Jul 2012 13:35:01 +0000 dan 2823@http://www.datastax.com/support-forums/ Hi there, I hope this is the right platform to raise this question. I want to set up a 3 Node Cluster with RF3. The data that i am holding isn't growing and is only used for read access, so I was thinking to have 3 nodes (on ec2, one in each availability zone), each holding the full dataset. My question is how many Seeds do I need? If I have one Seed and that Node goes down, how does the new node (the one that replaces the seed) figure out where to get the data from? Do I have to reconfigure the other nodes to become seeds? Ideally, I want to autoscale, so based on the amount of read requests, I want to be able to launch new nodes on demand. Sorry that might be a beginner questions, but I would be very grateful if someone could help me here. Thanks Dan joaquin on "aws ami" http://www.datastax.com/support-forums/topic/aws-ami#post-2596 Mon, 02 Jul 2012 21:05:52 +0000 joaquin 2596@http://www.datastax.com/support-forums/ I received a forwarded email about your concerns between the 2.1 and 2.2 AMIs. Both AMI's are nearly identical and most of the changes were small edge cases and a complete rewrite for cleaner code. But both are compatible on a higher level. Thanks, Joaquin joaquin on "aws ami" http://www.datastax.com/support-forums/topic/aws-ami#post-2589 Mon, 02 Jul 2012 19:25:14 +0000 joaquin 2589@http://www.datastax.com/support-forums/ Hello, This shouldn't have happened due to this line that ensures new versions will never been installed: <a href="https://github.com/riptano/ComboAMI/blob/2.2/ds2_configure.py#L743" rel="nofollow">https://github.com/riptano/ComboAMI/blob/2.2/ds2_configure.py#L743</a> To remove the node, simple run a <code>nodetool decommission</code> on the node. If this fails, you can kill the node and run a <code>nodetool removetoken</code> that will remove the entire token and any data that may be on the node, so use cautiously. For the full documentation on how to expand an AMI cluster, see: <a href="http://www.datastax.com/docs/1.1/install/expand_ami" rel="nofollow">http://www.datastax.com/docs/1.1/install/expand_ami</a> But do note that you will have to set the <code>--release</code> switch appropriately: <a href="https://github.com/riptano/ComboAMI/blob/2.2/SWITCHES.md" rel="nofollow">https://github.com/riptano/ComboAMI/blob/2.2/SWITCHES.md</a> Thanks, Joaquin denov on "aws ami" http://www.datastax.com/support-forums/topic/aws-ami#post-2588 Mon, 02 Jul 2012 18:57:34 +0000 denov 2588@http://www.datastax.com/support-forums/ somehow after a reboot of my aws datastax 2.1 instance i found that it upgraded cassandra from 1.0.7 to 1.1.1. how do i replace this node? it seem like the ami is only for setting up clusters and not for adding single nodes? thanks, deno