[Cloud Security] initializeCspIndices error handling

[opauloh]

After the recent changes the cloud_security_posture transform plugin, the initializeCspIndices function looks like this:

export const initializeCspIndices = async (esClient: ElasticsearchClient, logger: Logger) => {
  await Promise.allSettled([
    createPipelineIfNotExists(esClient, scorePipelineIngestConfig, logger),
    createPipelineIfNotExists(esClient, latestFindingsPipelineIngestConfig, logger),
  ]);

  const [
    createFindingsLatestIndexPromise,
    createVulnerabilitiesLatestIndexPromise,
    createBenchmarkScoreIndexPromise,
  ] = await Promise.allSettled([
    createLatestIndex(esClient, logger, latestIndexConfigs.findings),
    createLatestIndex(esClient, logger, latestIndexConfigs.vulnerabilities),
    createBenchmarkScoreIndex(esClient, logger),
  ]);

  if (createFindingsLatestIndexPromise.status === 'rejected') {
    logger.error(createFindingsLatestIndexPromise.reason);
  }
  if (createVulnerabilitiesLatestIndexPromise.status === 'rejected') {
    logger.error(createVulnerabilitiesLatestIndexPromise.reason);
  }
  if (createBenchmarkScoreIndexPromise.status === 'rejected') {
    logger.error(createBenchmarkScoreIndexPromise.reason);
  }
};

Now using Promise.allSettled we can handle each error independently, should we handle the rejected state of the createPipelineIfNotExists functions? Are the subsequent functions affected by the creation of the pipeline somehow?

Also, should we start collecting Telemetry data on these errors to know more about what circumstances makes them happen and how we can improve error handling in the future?

[elasticmachine]

Pinging @elastic/kibana-cloud-security-posture (Team:Cloud Security)

[CohenIdo]

Hey @opauloh, regarding having telemetry data on those errors: all our logs are automatically sent to centralize env, here you can find a dashboard the contains our logs.

Regarding the error handling, I agree, if one index failed it shouldn't impact the rest indices.

[opauloh]

Hey @opauloh, regarding having telemetry data on those errors: all our logs are automatically sent to centralize env, here you can find a dashboard the contains our logs.

That's great @CohenIdo, thanks