Is it possible to open connections to two buckets?

[n1rvana]

Looking at the source code, it looks like the answer is no. Not sure how libcouchbase handles it, but start_link will take a bucket, but if called again with another bucket it says "already open" or something like that.

Feel free to close this as wontfix if it's out the scope of what you need cberl for.

Mainly want to know if there's a way to do it and I'm missing it.

[wcummings]

You can create two separate connections (or connection pools, rather), each to a different bucket. I believe lcb does let you interact with multiple buckets from a single connection, cberl is just not structured this way right now.

1> cberl:start_link(default, 3, "localhost", "", "", "default").
{ok,<0.33.0>}
2> cberl:start_link(test, 3, "localhost", "", "", "test").
{ok,<0.40.0>}
3> cberl:set(test, <<"test">>, 0, <<"test">>).
ok
4> cberl:get(test, <<"test">>).
{<<"test">>,13776253455939141632,<<"test">>}
5> cberl:get(default, <<"test">>).
{<<"test">>,{error,key_enoent}}

The question is, are there situations where having multiple connections for multiple buckets is unacceptable? I've never run into this use-case, but it might be worth supporting.

[n1rvana]

PEBKC - I tried to use the same pool names. Separate pools is more than acceptable--ideal.

Also, you might find this interesting / useful:
http://trondn.blogspot.com/2013/08/connecting-to-multiple-clusters-from.html

[wcummings]

Supporting a per-operation bucket field is still a good idea. Marked as enhancement.