Loading...

Network: Robust Network Layer

Table of Contents

Overview

The SnapProtect Robust Network Layer improves reliability of backup and other operations by extending the window during which network connectivity may be temporarily lost, yet the job not be failed by the CommServe services that are managing it.

A CommCell environment consists of three primary components:

  • A CommServe host running CommServe services.
  • One or more MediaAgent hosts.
  • One or more client computers.

The CommServe server coordinates the activities within a CommCell environment. The CommServe server starts and manages backup jobs, receiving updates from them and recording their intermediate and final statuses. During normal operation, messages are exchanged between the CommServe server, the MediaAgent, and the clients. These messages indicate whether a backup job is still running.

Use Case

If the CommServe host loses network connectivity to the MediaAgent or client, the job may fail. A temporary loss of network connectivity (sometimes called a glitch) between the CommServe services and the MediaAgent and client is more common when remote CommServe host connectivity is through a WAN or VPN tunnel. Those arrangements are more complex and therefore more susceptible to network glitches. (Connectivity between MediaAgent and client is verified separately, through liveliness checks that run while a backup is in progress; see Status of a MediaAgent.)

In addition, with some agents, such as Exchange or Oracle, failing a job requires restarting the backup from the beginning. This could mean a substantial loss, in terms of time and used resources.

Note: Backup is only one operation that may be failed by lost network connectivity. Other operations can also be impacted by this scenario.

Solution

The SnapProtect Robust Network Layer protects the time and resources invested in a partially completed backup by delaying the failing of it. You determined the amount of delay by configuring the number of times, and the frequency at which the CommServe services attempt to contact the MediaAgent and client computers. The job is not failed until after this sequence has completed.

Support

The Robust Network Layer is supported on clients and MediaAgents running under Windows and UNIX.

Changing Robust Network Layer Configuration

During the data transfer phase of a backup operation, which runs on a MediaAgent and client, the Job Manager needs to periodically communicate directives to the CommServe instance. If during this time, a loss of network connectivity is detected, the Robust Network Layer enables the Job Manager to keep the job in the Running (Cannot be verified) job state, visible in the Job Controller, for a set period, while periodically checking for network connectivity, such as every 30 seconds for 20 minutes (these are the default values).

You can modify the Robust Network Layer configuration parameters as follows:

  1. From the CommCell console, expand Client Computers, then right-click the client  > Properties > Advanced.
  2. On the General tab, select the Enable retry on network errors checkbox.
    • In Retry Frequency, set the number of seconds at which the Job Manager checks network connectivity (default is 30 seconds).
    • In Retry Count, set the number of times the Job Manager checks for network connectivity (default is 40 times) before concluding that connectivity has been lost.
  3. Click OK.

Disabling Robust Network Layer

When a CommServe, Client or MediaAgent is installed, the Robust Network Layer is enabled by default.

Use these steps to disable Robust Network Layer:

  1. Open the CommCell console.
    1. To configure a client, expand Client Computers, then right-click the client  > Properties > Advanced.
    2. To configure a MediaAgent , expand Storage Resources > MediaAgents, then right-click the MediaAgent  > Properties and click the Control tab.
  2. Clear Enable retry on network errors to disnable the feature.
  3. Click OK.

You can re-enable the feature by selecting Enable retry on network errors.