Go library for managing a couchbase cluster via etcd and the couchbase REST API
had to write a little node script to process the json, but now it works ...
if you run the script, it'll create the full couchbase cluster, however, I still get to a setup screen when I launch the admin panel
and the second server isn't in the list either
is there more stuff in etcd required? or something I'm missing?
sorry, my fault! The etcd.pod.ip got lost
everything works like a charm now :)
thanks @tleyden !
thanks again for the PR's.
keep em' comin
If you want to update the sync gateway config for all your running SG nodes, is there a command / preferred way to do this? Looking in the cmd directory of this project I can see some commands around sync gateway configuration, but I'm not entirely sure how to use them or if I need to stop and remove the docker containers as described in http://tleyden.github.io/blog/2015/01/23/running-a-walrus-backed-sync-gateway-on-aws/
@tleyden do you actually need the etcd pod? I see that running in the GCE cluster when you bootstrap it
just wondering how I can share etcd info between clusters
as couchbase is running in one, and my app in another, don't know how to get to the db
Anybody tried https://github.com/couchbaselabs/couchbase-server-coreos on Windows Azure? I’ve gotten it to successfully work many times on AWS. I thought I’d give it a spin on Azure just for kicks. It works up to the point of finding nodes, but I’ve only found 2 of the 3 in my fleet after about two hours.
The nodes are in the same virtual network and cloud service so port 8091 should be available for the api check.
how do you boostrap data and views into a cluster btw.? Any recommendations?
I didn't know about that. So far I'm using a node.js bootstrap tool to insert all the views etc.
will try that, thanks!
How can I add a node to the cluster? I've X nodes running, and just want to add another, but I get this when running the $ docker run --ulimit nofile=40960:40960 --ulimit core=2000000:2000000 --ulimit memlock=2000000:2000000 --net=host derekdon/mm-couchbase-cluster-go update-wrapper couchbase-fleet launch-cbs --version 3.0.1 --num-nodes 4 --userpass "bla.bla" Tried num-nodes being 1, ie. add one, but I think it means total, so I set it to +1. Either way I get: Found residue -- key: /couchbase.com/couchbase-node-state in etcd. You should destroy the cluster first, then try again. I love and hate this setup. Everytime I want to make a change to something I seem to end up having to recreate my vps / start from scratch.
@derekdon I’m trying to get up to speed with docker for CBS and SG. Where are you running this docker run command?
I've several coreos vps (droplets) on digital ocean. I ssh into one of the database cluster nodes to run this command.
@jamiltz I've had a CBS and SG setup running successfully on this for months... had a lot of teething problems initially, but worked most of them out.