Skip to content

Latest commit

 

History

History
37 lines (30 loc) · 1.34 KB

disable_HA.md

File metadata and controls

37 lines (30 loc) · 1.34 KB

We can put together something for internal use as there are a series of commands that you can run through to understand the state of the platform.

To suspend HA, you can run the following command on the current active primary controller:

/opt/bluedata/bundles/bluedata*/startscript --action suspendha

Once you resolve any issues, then you can resume HA with the following command:

/opt/bluedata/bundles/bluedata*/startscript --action resumeha

Here is the basic process:

  1. From the primary controller, run bdconfig --getw and see if the hosts you expect to be primary and shadow are still designated as such or if they have switched (which would indicate a failover).

  2. Run this on primary, shadow and arbiter and send the output.

/opt/bluedata/common-install/bd_mgmt/bin/bd_mgmt ha query_state
  1. Run this on the primary controller and send the output:
/opt/bluedata/common-install/bd_mgmt/bin/bd_mgmt ha read_record bd_ha_record

4a. Run this on current primary and send the output:

pcs status

4b. If this shows an error like "Failed", run the following on primary and send the output: crm_resource -P5. Check if ha engine is running on primary and shadow ps auxf | grep ha_engine 11:46 There are a myriad of other questions that would also need to be collected. Things like: is the UI up and accessible? etc, etc.