Skip to content

Commit

Permalink
fix: Add documentation to bulkReadRows that each batch will process t… (
Browse files Browse the repository at this point in the history
#410)

* fix: Add documentation to bulkReadRows that each batch will process the keys sequentially

* fix: Add documentation to bulkReadRows that each batch will process the keys sequentially

* fix: Add documentation to bulkReadRows that each batch will process the keys sequentially
  • Loading branch information
dmitry-fa authored Sep 14, 2020
1 parent 2e17f0a commit 71dc8e3
Show file tree
Hide file tree
Showing 2 changed files with 16 additions and 7 deletions.
Original file line number Diff line number Diff line change
Expand Up @@ -76,7 +76,7 @@
* // One instance per application.
* BigtableTableAdminClient client = BigtableTableAdminClient.create("[PROJECT]", "[INSTANCE]");
*
* CreateTable request =
* CreateTableRequest request =
* CreateTableRequest.of("my-table")
* .addFamily("cf1")
* .addFamily("cf2", GCRULES.maxVersions(10))
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -57,7 +57,7 @@
* // One instance per application.
* BigtableDataClient client = BigtableDataClient.create("[PROJECT]", "[INSTANCE]")
*
* for(Row row : client.readRows(Query.create("[TABLE]")) {
* for(Row row : client.readRows(Query.create("[TABLE]"))) {
* // Do something with row
* }
*
Expand Down Expand Up @@ -1077,9 +1077,13 @@ public Batcher<RowMutationEntry, Void> newBulkMutationBatcher(@Nonnull String ta
}

/**
* Reads rows for given tableId in a batch. If the row does not exist, the value will be null.
* This operation should be called with in a single thread. The returned Batcher instance is not
* threadsafe, it can only be used from single thread.
* Reads rows for given tableId in a batch. If the row does not exist, the value will be null. The
* returned Batcher instance is not threadsafe, it can only be used from a single thread.
*
* <p>Performance notice: The ReadRows protocol requires that rows are sent in ascending key
* order, which means that the keys are processed sequentially on the server-side, so batching
* allows improving throughput but not latency. Lower latencies can be achieved by sending smaller
* requests concurrently.
*
* <p>Sample Code:
*
Expand Down Expand Up @@ -1113,8 +1117,13 @@ public Batcher<ByteString, Row> newBulkReadRowsBatcher(String tableId) {

/**
* Reads rows for given tableId and filter criteria in a batch. If the row does not exist, the
* value will be null. This operation should be called with in a single thread. The returned
* Batcher instance is not threadsafe, it can only be used from single thread.
* value will be null. The returned Batcher instance is not threadsafe, it can only be used from a
* single thread.
*
* <p>Performance notice: The ReadRows protocol requires that rows are sent in ascending key
* order, which means that the keys are processed sequentially on the server-side, so batching
* allows improving throughput but not latency. Lower latencies can be achieved by sending smaller
* requests concurrently.
*
* <p>Sample Code:
*
Expand Down

0 comments on commit 71dc8e3

Please sign in to comment.