Fix bug in calculating when ad-hoc tasks are out of attempts

[mikecote]

In this PR, I'm fixing a bug where ad-hoc tasks would have one fewer attempts to retry in failure scenarios when using mget.

To verify

1. Apply the following diff to your code

diff --git a/x-pack/plugins/stack_connectors/server/connector_types/server_log/index.ts b/x-pack/plugins/stack_connectors/server/connector_types/server_log/index.ts
index 0275b2bdc2f..d481c3820a1 100644
--- a/x-pack/plugins/stack_connectors/server/connector_types/server_log/index.ts
+++ b/x-pack/plugins/stack_connectors/server/connector_types/server_log/index.ts
@@ -77,6 +77,10 @@ export function getConnectorType(): ServerLogConnectorType {
 async function executor(
   execOptions: ServerLogConnectorTypeExecutorOptions
 ): Promise<ConnectorTypeExecutorResult<void>> {
+
+  console.log('*** Server log execution');
+  throw new Error('Fail');
+
   const { actionId, params, logger } = execOptions;

   const sanitizedMessage = withoutControlCharacters(params.message);
diff --git a/x-pack/plugins/task_manager/server/config.ts b/x-pack/plugins/task_manager/server/config.ts
index db07494ef4f..07e277f8d16 100644
--- a/x-pack/plugins/task_manager/server/config.ts
+++ b/x-pack/plugins/task_manager/server/config.ts
@@ -202,7 +202,7 @@ export const configSchema = schema.object(
       max: 100,
       min: 1,
     }),
-    claim_strategy: schema.string({ defaultValue: CLAIM_STRATEGY_UPDATE_BY_QUERY }),
+    claim_strategy: schema.string({ defaultValue: CLAIM_STRATEGY_MGET }),
     request_timeouts: requestTimeoutsConfig,
   },
   {
diff --git a/x-pack/plugins/task_manager/server/lib/get_retry_at.ts b/x-pack/plugins/task_manager/server/lib/get_retry_at.ts
index 278ba18642d..c8fb911d500 100644
--- a/x-pack/plugins/task_manager/server/lib/get_retry_at.ts
+++ b/x-pack/plugins/task_manager/server/lib/get_retry_at.ts
@@ -54,6 +54,7 @@ export function getRetryDate({
 }

 export function calculateDelayBasedOnAttempts(attempts: number) {
+  return 10 * 1000;
   // Return 30s for the first retry attempt
   if (attempts === 1) {
     return 30 * 1000;

2. Create an always firing rule that runs every hour, triggering a server log on check intervals
3. Let the rule run and observe the server log action running and failing three times (as compared to two)

[elasticmachine]

Pinging @elastic/response-ops (Team:ResponseOps)

[kibana-ci]

💛 Build succeeded, but was flaky

• Buildkite Build
• Commit: adb5f29

Failed CI Steps

Metrics [docs]

✅ unchanged

To update your PR or re-run it, just comment with:
@elasticmachine merge upstream

cc @mikecote

[kibanamachine]

💔 Some backports could not be created

Status	Branch	Result
❌	8.16	The branch "8.16" does not exist
✅	8.x

Note: Successful backport PRs will be merged automatically after passing CI.

Manual backport

To create the backport manually run:

node scripts/backport --pr 192907

Questions ?

Please refer to the Backport tool documentation