Reset Search



8224 - Virtual Traffic Manager (vTM) web UI displays "Cluster: Conflict" message briefly, then switches back to green OK status

« Go Back


Last Modified Date12/8/2017 8:00 PM
This article describes an issue where a "Cluster: Conflict" message briefly appears in the VTM Web UI then switches back to green OK status.

Problem or Goal
This issue can occur under the following conditions:
  • Traffic Manager is clustered with a pool node that is a backend Amazon Web Services (AWS) Simple Storage Server (S3) bucket.
  • The pool has been configured using DNS-derived auto-scaling to ensure it uses the correct S3 IP address.
  • The S3 IP address changes periodically.
If an HTTP request is being serviced at the time that the DNS-derived autoscaling is updates, the HTTP response is truncated and the end-user only receives part of the file that is downloaded.
With DNS-derived autoscaling, the pool configuration gets updated regularly.  During these regular updates, the node is removed from the pool configuration briefly, then is rejoined to the pool.  This results in the "Cluster: Conflict" message being displayed in the VTM Web UI briefly.  Additionally, if the node is servicing a connection at the time that an update is performed, then the connection is immediately terminated by default which can result in truncated HTTP responses and incomplete file downloads.
The default setting for DNS-derived autoscaling will cause the node to be removed from the pool up to four times a minute.  To change the default behavior and resolve this issue do the following:

Go to Pool > Connection Management > node_delete_behavior and change the default setting from:

All connections to node are closed immediately


Allow existing connections to the node to finish before deletion
Related Links
Attachment 1 
Created ByVenkataKondaReddy Palem



Was this article helpful?



Please tell us how we can make this article more useful.

Characters Remaining: 255