[
https://issues.apache.org/jira/browse/PHOENIX-1226?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Jesse Yates updated PHOENIX-1226:
---------------------------------
Attachment: phoenix-1226-4.0-v0.patch
Adding patch that just assumes all the values are written as Bytes.toBytes()
from Strings. This is actually how we are storing them anyways through
TracingCompat, so it should be fine for the moment.
Ideally, we'd actually like to store each of the kv annotations as the
underlying bytes, and then let the client decide how it wants to deserialize
them. However, I think that also requires a schema upgrade of the tags column
(assumes its string everywhere) and there is a bug in the blob array
implementation in phoenix around zero-byte elements (PHOENIX-1223). Plus it
will require significantly updating the underlying transport, which currently
assumes everything is a MetricsTag (and stores the value as a string)
That isn't an immediate problem here because everything is strings, but if we
want to make it a general catch-all trace sink, that needs to be fixed first.
> Exception in Tracing
> --------------------
>
> Key: PHOENIX-1226
> URL: https://issues.apache.org/jira/browse/PHOENIX-1226
> Project: Phoenix
> Issue Type: Bug
> Affects Versions: 4.1
> Environment: 0.98.5 hbase, 4.1.0 phoenix
> Reporter: Dan Di Spaltro
> Assignee: Jesse Yates
> Attachments: phoenix-1226-4.0-v0.patch
>
>
> I was exposed to an exception in the tracing code, during my test setup of
> Phoenix in the following code:
> {code}
> 58062 [defaultRpcServer.handler=2,queue=0,port=53950] WARN
> org.apache.hadoop.ipc.RpcServer -
> defaultRpcServer.handler=2,queue=0,port=53950: caught:
> java.lang.IllegalArgumentException: offset (0) + length (4) exceed the
> capacity of the array: 3
> at
> org.apache.hadoop.hbase.util.Bytes.explainWrongLengthOrOffset(Bytes.java:600)
> at org.apache.hadoop.hbase.util.Bytes.toInt(Bytes.java:749)
> at org.apache.hadoop.hbase.util.Bytes.toInt(Bytes.java:725)
> at
> org.apache.phoenix.trace.TracingCompat.readAnnotation(TracingCompat.java:56)
> at
> org.apache.phoenix.trace.TraceMetricSource.receiveSpan(TraceMetricSource.java:121)
> at org.cloudera.htrace.Tracer.deliver(Tracer.java:81)
> at org.cloudera.htrace.impl.MilliSpan.stop(MilliSpan.java:70)
> at org.cloudera.htrace.TraceScope.close(TraceScope.java:70)
> at org.apache.hadoop.hbase.ipc.CallRunner.run(CallRunner.java:106)
> at
> org.apache.hadoop.hbase.ipc.RpcExecutor.consumerLoop(RpcExecutor.java:114)
> at org.apache.hadoop.hbase.ipc.RpcExecutor$1.run(RpcExecutor.java:94)
> at java.lang.Thread.run(Thread.java:744)
> {code}
> It is related to the following line of code where we interpret all KV
> annotation values as byte-wise integers here:
> https://github.com/apache/phoenix/blob/v4.1.0/phoenix-hadoop-compat/src/main/java/org/apache/phoenix/trace/TracingCompat.java#L56
> Here is where HBase is adding a non-integer KV annotation:
> https://github.com/apache/hbase/blob/0.98.5/hbase-server/src/main/java/org/apache/hadoop/hbase/ipc/RequestContext.java#L105
> The fix should be simple, but I am not aware of all the related issues in
> changing this.
> cc [~jesse_yates], [[email protected]], [~giacomotaylor]
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
