pg-queue

Database based queue system

Module specifics

ee-diags

Application deploy needs this:

    "hazelcast": {
        "clusterName": "${HAZELCAST_CLUSTER}",
        "mapConfig": {
            "known_queues": {
                "ttl": "PT72H",
                "maxIdle": "PT24H",
                "sizePerNode": 1000,
                "format": "BINARY",
                "backupCount": 0,
                "asyncBackupCount":  0,
                "nearCache": {
                    "ttl": "PT24H",
                    "maxIdle": "PT24H",
                    "sizePerNode": 100,
                    "format": "OBJECT"
                }
            }
        }
    }

And a headless dns name for the service... and payara network-policies.

Configuration is in: src/main/resources/pg-queue-endpoint-config.json

{
    "dataSource": "jdbc/<datasource>",
    "jobLogMapper": "<package>.<class>.<static-instance of JobLogMapper type>",
    "systemName": "<service> - queue admin",
    "diagPercentMatch": <collapsing queue error match ie 70>
}

Throttle

Throttle syntax is a comma separated list of: /[]TIMEUNIT[!]

• number of failures for throttle to go into effect
• in a given period (defaults to 1)
  • defaults to 1
  • takes one of
    • ms
    • s
    • m
    • h
• should this throttle persist even if a successful result is registered

eq. thread rule, when job starts and job status

• 1/2s!,5/m
  • 0s fail
  • 2s fail
  • 4s fail
  • 6s fail
  • 8s fail
  • 60s fail
  • 62s fail
• 1/2s!,5/m
  • 0s fail
  • 2s success
  • 2s fail
  • 4s fail
• 1/2s!
  • 0s fail
  • 2s success
  • 2s fail
  • 4s fail
• 1/2s!
  • 0s fail
  • 1s success (other thread)
  • 2s fail
  • 4s fail
• 1/2s
  • 0s fail
  • 1s success (other thread)
  • 1s fail
  • 3s fail