For a wider audience, it would be interesting to do an implementation of Erlang
distribution based on HTTP.
I've been contemplating that to run Erjang apps on Google AppEngine. But it
wouldn't solve the issue that a load balancer in front of such a cluster of
HTTP enabled Erlang nodes will make
Getting dynos to communicate might be tricky. When you make an HTTP
request to heroku, the request first hits the heroku "routing mesh". The
mesh is responsible for deciding which dyno will service the request. I
don't know of any way to explicitly send a request to a specific dyno.
Dynos are al
So long as the dynos can communicate with one another and cooperate via node
communication, the stateless nature of the application doesn't really matter. I
don't need to persist anything once a dyno is shut down.
The thing that makes me curious is whether the things riak_core (not riak_kv)
de
Hi Jeff,
Dynos on Heroku don't have access to any permanent disk storage. They can
read and write to /tmp, but that is not guaranteed to stick around.
Dynos are also restricted to only accept HTTP connections. This would rule
out any ProtocolBuffer connections.
> Curious if anyone is able to
Curious if anyone is able to use clustered nodes on Heroku using the Erlang
buildpack [1]? Seems like it should at least be theoretically possible to
launch new dynos that become part of your node cluster. How reliable/easy/hard
is node communication on EC2?
[1] - https://github.com/heroku/hero
Hi Mark,
I resized the cluster from 4x1GB RAM to 4x4GB RAM. Also increased
{map_js_vm_count, 8 } to {map_js_vm_count, 48 } and {reduce_js_vm_count,
6 } to {reduce_js_vm_count, 36 } inside app.config but still have the
same problem from time to time...
The function I use to do the link-walki
Hi Mårten,
We've migrated our issue tracking to GitHub issues, so most bugs have been
moved to there. This particular bug represents a feature request- we've
migrated those to another tool that is currently only visible to Basho
employees. Some of the content there probably should be visible
Honk honk,
I've been watching the "Delete all keys in a bucket"[1] issue which
just got closed with the status "CLOSED MIGRATED". What does this
status indicate?
cheers, m.
[1] https://issues.basho.com/show_bug.cgi?id=79
___
riak-users mailing list
Thanks guys,
I'm more inclined to have an API like get_original_metadata and
get_metadata. The get_metadata in this case always returns whatever
metadata is set on the object, new or original. In the current API, if
calling get_update_metadata will return the original metadata if there are
no ch
