Site Recovery Manager, vSphere Replication & Disaster Recovery as a Service Fails to connect to remote site - Operation timed out: 300 seconds

Products

• VMware Aria Suite
• VMware Live Recovery

Issue/Introduction

Symptoms:

Connecting to the remote site within Site Recovery Manager/vSphere Replication or DRaaS counter parts may result in a connection time stating the server is down, after trying for 300 seconds.

You may see errors like:

Caused by: com.vmware.vim.vmomi.client.exception.ConnectionException: https://remote_site_ip:8043 invocation failed with "java.net.SocketTimeoutException"
Caused by: java.net.SocketTimeoutException

Cause

A common issue that may cause this is the MTU settings between the SRM/VR server with the remote site being uniform. This article covers this cause.

Resolution

Begin by checking your MTU configuration from the source DR solution and the destination site. If you find that MTU setting is higher than 1500 on some connecting network parts, you will need to uniform this configuration. By default MTU is 1500, however some customers may use jumbo frames with an MTU size of 9000. 

It is important to note that even though the MTU size is set to 1500 or 9000 in the example above, there is overhead that must be considered. For example, packet overhead for MTU of 9000 equates to a payload size of 8972. When testing pings with MTU it is important to consider this.

Example: ping <target IP/DNS address> -s 8972

Our DR solutions require that  MTU uniformity is applied from start to end.

Troubleshooting this may also require you to allow ICMP packets if this is not allowed in your configuration.