[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: gear creation on nodes



Ah.. That's what I was thinking.
Thx

Sent from my iPhone

On Jan 28, 2013, at 6:04 PM, Krishna Raman <kraman gmail com> wrote:

Id you are in a non-production environment, then you can turned districts off.
In that case, as long as mcollective is configured properly and the DNS entry for the node is correct, it will be picked up and gears will be created on there.

--Krishna

On Jan 28, 2013, at 6:01 PM, Meghdoot wrote:

Many thanks.
Ok so adding to the district is sort of explicit registration of the node. I was thinking districts are optional say in non production settings and nodes got registered through some mcollective query. 

Sent from my iPhone

On Jan 28, 2013, at 5:45 PM, Krishna Raman <kraman gmail com> wrote:

Hi,

That is the main function responsible for choosing the location for the next gear. Grant covered most of the algorithm. The algorithm also tries to also find a node which does not host 
any other gears from application.

After creating a new node, you need to assign it to an existing or new district, at which point, it is included in the selection algorithm.

If a node goes down, then it will not respond to the n-collective query and so it is automatically taken out of the pool of possible nodes. However, any gears that were already on that node
are not automatically migrated.

--Krishna

On Jan 28, 2013, at 5:28 PM, Grant Shipley wrote:

Algorithm for finding an available node

• Find the district with the most available UUIDs.
• Find the nodes within that district that have the least active_capacity.
• If some nodes still have active capacity, randomly pick one of the nodes with the lower levels of active_capacity.
• If no nodes in that district have active capacity, find nodes within any district (with available UUIDs) that have available capacity and randomly pick one of the nodes with the lower levels of active_capacity.
This algorithm means that applications are created on the districts with the most available UUIDs first, and districts that have available, non-active capacity are preferred.


--
gs
On Jan 28, 2013, at 7:40 PM, meghdoot bhattacharya <meghdoot_b yahoo com> wrote:

Hi,
        Given a set of nodes, when new gears are about to be created, what is the algorithm used to select the node?

        If brand new nodes/VM's are added later, how do those new nodes get discovered and made available to haproxy for future gear creation on those nodes? Does the mcollective in broker periodically send messages to discover new nodes or figure out if nodes are gone?

Thx.
_______________________________________________
dev mailing list
dev lists openshift redhat com
http://lists.openshift.redhat.com/openshiftmm/listinfo/dev

_______________________________________________
dev mailing list
dev lists openshift redhat com
http://lists.openshift.redhat.com/openshiftmm/listinfo/dev



[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]