usage.txt

Usage: spiegel --type=type --url=couchdb-url options

  --type=type             The type of the process: update-listener, change-listener, replicator,
                          install or uninstall. The install and uninstall processes run and then
                          exit, but the update-listener, change-listener and replicator will run
                          indefinitely. You can run as many update-listeners, change-listeners and
                          replicators as your DB setup can handle. In most cases you'll want to run
                          at least 2 of each of these processes for redundancy. In general, if you
                          need to listen to more changes or respond to these changes faster, add a
                          change-listener. Similarly, if you need to perform more replications or
                          replicate faster, add a replicator.

                          install           Creates the spiegel DB and sieve

                          uninstall         Destroys the spiegel DB and sieve

                          update-listener   The update-listener listens for all DB updates and then
                                            schedules on_changes and replications accordingly.

                          change-listener   The change-listener runs on_change rules for all
                                            matching changes

                          replicator        The replicator performs replications

  --url=couchdb-url       The URL to the CouchDB instance

Options:

  --log-level=level         error|warn|info|debug. Setting a particular level implies that all log
                            records at that level and above are logged.

                            Default: info

  --db-name=name            The name of the spiegel DB.

                            Default: spiegel

  --namespace=namespace     The namespace used to prefix names of global docs, e.g. the sieve in
                            _global_changes. You can use this namespace to keep test data separate
                            from other data.

                            Default: ''

  --batch-timeout=ms        The time in milliseconds to wait after an update before the batch is
                            considered done regardless of whether there are any more updates.

                            Applies to: update-listener
                            Default: 1000

  --save-seq-after=secs     The time in seconds after which the last _global_changes seq number is
                            saved. This seq number is used to avoid re-processing updates whenever
                            an update-listener is started or restarted.

                            Applies to: update-listener
                            Default: 60

  --passwords-file=path     A JSON file with passwords for hostnames and usernames, e.g.
                            {
                              "example.com": {
                                "user1": "secret1",
                                "user2": "secret2"
                              },
                              "google.com": {
                                "larry": "mypassword"
                              }
                            }

                            Applies to: change-listener or replicator

  --batch-size=size         The maximum number of items that will be processed in a batch.

                            Applies to: update-listener, change-listener
                            Default: 100

  --retry-after=secs        A failed item will be reprocessed after this many seconds. Warning: secs
                            must be less than the maximum time it takes to perform the action or
                            else there can be concurrent actions for the same DB that will backup
                            the queue and continuously run.

                            Applies to: change-listener, replicator
                            Default: 10800

  --concurrency=num         The maximum number of items that will be processed concurrently.
                            Warning: setting this value too high can result in the process using a
                            lot of memory as it queues tasks.

                            Applies to: change-listener, replicator
                            Default: 20

  --check-stalled=secs      It takes check-stalled + retry-after secs before an action is retried.
                            Be careful not make check-stalled too low though or else you'll waste a
                            lot of CPU cycles just checking for stalled processes.

                            Applies to: change-listener, replicator
                            Default: 600

  --version                 Output the version