Your Browser is not longer supported

Please use Google Chrome, Mozilla Firefox or Microsoft Edge to view the page correctly
Loading...

{{viewport.spaceProperty.prod}}

High Availability (HA)

&pagelevel(5)&pagelevel

ROBAR-SV supports High Availability (HA) configurations for SE Servers as described below.
Archive system ETERNUS CS (ABBA interface) can be controlled via 2 IP ports. This allows ROBAR-SV to operate it with 2 instances running simultaneously on 2 different Management Units in HA manner.
These HA instances should differ only in the archive system IP port in the robot_address1 parameter but also may differ in the device configuration (see below).

To support HA for ROBAR-SV, currently there are 2 configuration options provided, their pros and cons are detailed below.

HA configuration with independent device sets

This configuration requires each instance to have an independent device set (none of the devices can be shared between instances). Device sets can be unequal and contain different number of devices depending on usage requirements.
Device set separation between instances allows ROBAR-SV to maintain correct device status tracking when processing mount/unmount requests.

ProsCons
Continuously consistent device status including during mount/unmount requests without AMU errors NxxAdditional devices are required

Instances have different device configurations

HA configuration with shared device sets

In this configuration devices are shared between 2 instances. It may cause inconsistencies with device status tracking in ROBAR-SV because the archive system reports the status independently to each instance upon processing its request. Such inconsistencies may include:

  • reporting mounted volume on a device which is unmounted by another instance and already free
  • reporting device as free when it is in fact occupied with a volume mounted by another instance

To operate in this HA configuration, ROBAR-SV provides an additional parameter IGNORE_DEV_STATUS=YES (default: NO) to be defined in the gen_env_vars section of each instance's configuration. In this case ROBAR-SV ignores internal device status tracking and relies only on information returned by the archive system which may report AMU errors like N202, N203, N303, N307, N309, etc.

With the setting IGNORE_DEV_STATUS=NO, ROBAR-SV uses internal device status tracking.

ProsCons
No need for additional devices, one set of devices is sharedPotentially inconsistent device status on the 2 instances during mount/unmount requests for the same device
Almost the same instance configuration except for the archive system IP port