HBASE-6449 added support for tracing requests through HBase, using the open source tracing library, HTrace. Setting up tracing is quite simple, however it currently requires some very minor changes to your client code (it would not be very difficult to remove this requirement).
The tracing system works by collecting information in structs called 'Spans'. It is up to
you to choose how you want to receive this information by implementing the
SpanReceiver
interface, which defines one method:
public void receiveSpan(Span span);
This method serves as a callback whenever a span is completed. HTrace allows you to use as many SpanReceivers as you want so you can easily send trace information to multiple destinations.
Configure what SpanReceivers you'd like to us by putting a comma separated list of the
fully-qualified class name of classes implementing SpanReceiver
in
hbase-site.xml
property:
hbase.trace.spanreceiver.classes
.
HTrace includes a LocalFileSpanReceiver
that writes all span
information to local files in a JSON-based format. The
LocalFileSpanReceiver
looks in hbase-site.xml
for a hbase.local-file-span-receiver.path
property with a value describing
the name of the file to which nodes should write their span information.
<property> <name>hbase.trace.spanreceiver.classes</name> <value>org.htrace.impl.LocalFileSpanReceiver</value> </property> <property> <name>hbase.local-file-span-receiver.path</name> <value>/var/log/hbase/htrace.out</value> </property>
HTrace also provides ZipkinSpanReceiver
which converts spans to Zipkin span format and send them to
Zipkin server. In order to use this span receiver, you need to install the jar of
htrace-zipkin to your HBase's classpath on all of the nodes in your cluster.
htrace-zipkin
is published to the maven central repository. You could get
the latest version from there or just build it locally and then copy it out to all nodes,
change your config to use zipkin receiver, distribute the new configuration and then (rolling)
restart.
Here is the example of manual setup procedure.
$ git clone https://github.com/cloudera/htrace $ cd htrace/htrace-zipkin $ mvn compile assembly:single $ cp target/htrace-zipkin-*-jar-with-dependencies.jar $HBASE_HOME/lib/ # copy jar to all nodes...
The ZipkinSpanReceiver
looks in hbase-site.xml
for a hbase.zipkin.collector-hostname
and
hbase.zipkin.collector-port
property with a value describing the Zipkin
collector server to which span information are sent.
<property> <name>hbase.trace.spanreceiver.classes</name> <value>org.htrace.impl.ZipkinSpanReceiver</value> </property> <property> <name>hbase.zipkin.collector-hostname</name> <value>localhost</value> </property> <property> <name>hbase.zipkin.collector-port</name> <value>9410</value> </property>
If you do not want to use the included span receivers, you are encouraged to write your
own receiver (take a look at LocalFileSpanReceiver
for an example). If
you think others would benefit from your receiver, file a JIRA or send a pull request to HTrace.
In order to turn on tracing in your client code, you must initialize the module sending spans to receiver once per client process.
private SpanReceiverHost spanReceiverHost; ... Configuration conf = HBaseConfiguration.create(); SpanReceiverHost spanReceiverHost = SpanReceiverHost.getInstance(conf);
Then you simply start tracing span before requests you think are interesting, and close it when the request is done. For example, if you wanted to trace all of your get operations, you change this:
HTable table = new HTable(conf, "t1"); Get get = new Get(Bytes.toBytes("r1")); Result res = table.get(get);
into:
TraceScope ts = Trace.startSpan("Gets", Sampler.ALWAYS); try { HTable table = new HTable(conf, "t1"); Get get = new Get(Bytes.toBytes("r1")); Result res = table.get(get); } finally { ts.close(); }
If you wanted to trace half of your 'get' operations, you would pass in:
new ProbabilitySampler(0.5)
in lieu of Sampler.ALWAYS
to Trace.startSpan()
.
See the HTrace README
for more information on Samplers.
You can use trace command for tracing requests from HBase Shell. trace 'start' command turns on tracing and trace 'stop' command turns off tracing.
hbase(main):001:0> trace 'start' hbase(main):002:0> put 'test', 'row1', 'f:', 'val1' # traced commands hbase(main):003:0> trace 'stop'
trace 'start' and trace 'stop' always returns boolean value representing if or not there is ongoing tracing. As a result, trace 'stop' returns false on suceess. trace 'status' just returns if or not tracing is turned on.
hbase(main):001:0> trace 'start' => true hbase(main):002:0> trace 'status' => true hbase(main):003:0> trace 'stop' => false hbase(main):004:0> trace 'status' => false