Age | Commit message (Collapse) | Author | Files | Lines |
|
By default, OpenStack will schedule you into the right availablity
zone that has capacity, hard coding this can potentially result in
not being able to provision new virtual machines.
This patch drops the hard-wired value and lets Nova on the
scheduling size pick the correct availability zone.
Signed-off-by: Mohammed Naser <mnaser@vexxhost.com>
Change-Id: Id87111f06ad702daedb947f7e4d67f4cbbc2ad1a
|
|
Change-Id: Ia8bdd474fcc1477080b6b44bc80600808b380335
Signed-off-by: C.J. Collier <cjcollier@linuxfoundation.org>
|
|
Switch to Interlink1-6 networks for multilink labs
Change-Id: I119a9751e96ebf787e23234cc53a5882b419976b
Signed-off-by: Andrew Grimberg <agrimberg@linuxfoundation.org>
|
|
When nodepool brings online a node and subnodes they are presently all
named the same way. This makes it difficult to pinpoint the leader node
of a set if we need to do any work directly against OpenStack. This
change will cause sub-nodes to have the prefix of np-sub- and the rest
of the name from the leader node making it clear.
Change-Id: I536b2e4437a93f118436e8eeb6869a2f691be0f6
Signed-off-by: Andrew Grimberg <agrimberg@linuxfoundation.org>
|
|
This adds in the nodepool configuration that we are presently using.
This is the starting point for the community to help define what build
slave definitions are needed.
Change-Id: I1c6325869a0a411265994090f9dc754de03ea173
Signed-off-by: Andrew Grimberg <agrimberg@linuxfoundation.org>
|