summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorAlexander Shorin <kxepal@apache.org>2013-12-14 15:37:13 +0400
committerAlexander Shorin <kxepal@apache.org>2013-12-14 15:38:19 +0400
commitcb9095053b3f5bddaae9dae9721d4d29a01734f9 (patch)
tree4a50cef170701e81ca8a95ed2a6ba102bb449b72
parent5e2a4278ea50d284c75469a4cf3c5a2d26ceb63d (diff)
downloadcouchdb-cb9095053b3f5bddaae9dae9721d4d29a01734f9.tar.gz
Document use_checkpoints and checkpoint_interval config options.
-rw-r--r--share/doc/src/config/replicator.rst37
-rw-r--r--share/doc/src/json-structure.rst2
-rw-r--r--share/doc/src/replication/replicator.rst4
3 files changed, 41 insertions, 2 deletions
diff --git a/share/doc/src/config/replicator.rst b/share/doc/src/config/replicator.rst
index 45451c1f8..515c6797f 100644
--- a/share/doc/src/config/replicator.rst
+++ b/share/doc/src/config/replicator.rst
@@ -111,6 +111,43 @@ Replicator Database Configuration
.. _inet: http://www.erlang.org/doc/man/inet.html#setopts-2
+ .. config:option:: checkpoint_interval
+
+ .. versionadded:: 1.6
+
+ Defines replication checkpoint interval in milliseconds. :ref:`Replicator
+ <replication/replicator>` will :get:`requests </{db}>` Source database
+ each time after a specified amount of interval::
+
+ [replicator]
+ checkpoint_interval = 5000
+
+ Keeping this value lower raises polling rate to Source database, which
+ may be useful for frequently changing data, while higher value saves
+ traffic usage that is important for mobile clients.
+
+
+ .. config:option:: use_checkpoints
+
+ .. versionadded:: 1.6
+
+ If set as ``true`` CouchDB makes intermediate checkpoints during
+ replication from those it may continue it in case of failure or restart::
+
+ [replicator]
+ use_checkpoints = true
+
+ .. warning:: It's **not recommended** to switch this option off
+ (by setting value ``false`` to it) since without checkpoints there would
+ not be stored any replication history logs both on Source and Target,
+ forcing replication to start from the very beginning.
+
+ .. note:: Checkpoints (aka Replication Log) are stored as :ref:`local
+ documents <api/local>` both on Source and Target endpoints. In case,
+ when client has no write permission on Source database you may disable
+ them, but locally for specific replication process.
+ See :ref:`replication-settings` for more info.
+
.. config:option:: cert_file
diff --git a/share/doc/src/json-structure.rst b/share/doc/src/json-structure.rst
index 042df59e9..7d22f1e9d 100644
--- a/share/doc/src/json-structure.rst
+++ b/share/doc/src/json-structure.rst
@@ -274,6 +274,8 @@ Replication Settings
| use_checkpoints (optional) | Whether to use replication checkpoints |
| | or not |
+--------------------------------+---------------------------------------------+
+| checkpoint_interval (optional) | Specifies checkpoint interval in ms. |
++--------------------------------+---------------------------------------------+
.. _replication-status:
diff --git a/share/doc/src/replication/replicator.rst b/share/doc/src/replication/replicator.rst
index e93c5791f..5117752ff 100644
--- a/share/doc/src/replication/replicator.rst
+++ b/share/doc/src/replication/replicator.rst
@@ -22,8 +22,8 @@ cancel an ongoing replication.
These documents have exactly the same content as the JSON objects we use to
``POST`` to ``_replicate`` (fields ``source``, ``target``, ``create_target``,
-``continuous``, ``doc_ids``, ``filter``, ``query_params``,
-``use_checkpoints``).
+``continuous``, ``doc_ids``, ``filter``, ``query_params``, ``use_checkpoints``,
+``checkpoint_interval``).
Replication documents can have a user defined ``_id`` (handy for finding a
specific replication request later). Design Documents