[ARVADOS] updated: cb7c9d8d80ec2bd9a42a9e5b4b9bfc5fed6da1a1

[git at public.curoverse.com]

Summary of changes:
 services/nodemanager/arvnodeman/computenode/dispatch/__init__.py | 8 ++++++++
 1 file changed, 8 insertions(+)

       via  cb7c9d8d80ec2bd9a42a9e5b4b9bfc5fed6da1a1 (commit)
      from  cd1109efb038bf5ad1991f6c0a185ae570edba73 (commit)

Those revisions listed above that are new to this repository have
not appeared on any other notification email; so we list those
revisions in full, below.


commit cb7c9d8d80ec2bd9a42a9e5b4b9bfc5fed6da1a1
Author: Brett Smith <brett at curoverse.com>
Date:   Mon Mar 16 10:23:27 2015 -0400

    5313: Fixup comment to explain why update_arvados_node doesn't retry.

diff --git a/services/nodemanager/arvnodeman/computenode/dispatch/__init__.py b/services/nodemanager/arvnodeman/computenode/dispatch/__init__.py
index 047f23b..7081762 100644
--- a/services/nodemanager/arvnodeman/computenode/dispatch/__init__.py
+++ b/services/nodemanager/arvnodeman/computenode/dispatch/__init__.py
@@ -340,6 +340,14 @@ class ComputeNodeMonitorActor(config.actor_class):
             self._later.consider_shutdown()
 
     def update_arvados_node(self, arvados_node):
+        # If the cloud node's FQDN doesn't match what's in the Arvados node
+        # record, make them match.
+        # This method is a little unusual in the way it just fires off the
+        # request without checking the result or retrying errors.  That's
+        # because this update happens every time we reload the Arvados node
+        # list: if a previous sync attempt failed, we'll see that the names
+        # are out of sync and just try again.  ComputeNodeUpdateActor has
+        # the logic to throttle those effective retries when there's trouble.
         if arvados_node is not None:
             self.arvados_node = arvados_node
             if (self._cloud_node_fqdn(self.cloud_node) !=

-----------------------------------------------------------------------


hooks/post-receive
--