The SAS® Event Stream Processing Standalone Installer update fails with a "work directory check failed" error


The SAS Event Stream Processing Standalone Installer (ESI) update flow fails with the following error message:

sas-esp-esi.sh workflow --name=update run

work directory check failed: expected cluster resources to exist at "/workspace/deploy/aio-cluster.yaml", but it does not

or

work directory check failed: expected base resources to exist at \"/workspace/deploy/base\", but it does not

This error occurs when you run the sas-esp-esi.sh workflow --name=update command after a previous failure of an update command.

For example, an issue occurred with accessing the SAS® Viya® Orders API, which caused a previous update to fail. The sas-esp-esi.sh workflow --name=update command moved assets in the deploy directory to prepare for the update. When the update command failed, these assets were not restored. 

Resolution or Workaround

As a workaround, you can restore the assets manually by running the following commands

  1. cd $SAS_ESP_ESI_WORK_DIR
  2. cp last-applied/aio* .
  3. cp -r last-applied/base .
  4. cp last-applied/license.jwt ../order 

The sas-esp-esi.sh workflow --name=update command no longer fails with the work directory check failed... message. If the command fails for another reason, run the workaround commands again to restore the assets.

A future version of ESI will restore these assets automatically.