A lock on the job — no double runs

The scenario sets a "busy" mark in Redis for the duration of a long job: a second run sees it, stops with a clear error, and leaves the data intact.

• Only one run at a time
• No duplicated data
• The second run fails clearly, not silently
• The lock expires — a hung job can't block everything

RedisAirtableAWS DynamoDBAzure Cosmos DBBaserowData tableGristMicrosoft SharePointMonday.comMonica CRMNocoDBSeaTableServiceNowStrapiSupabaseWebflowWekan