drill-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "abdelhakim deneche" <adene...@gmail.com>
Subject Re: Review Request 32987: DRILL-2675: Implement a subset of User Exceptions to improve how errors are reported to the user
Date Thu, 09 Apr 2015 19:12:32 GMT


> On April 9, 2015, 6:05 p.m., Jacques Nadeau wrote:
> > common/src/main/java/org/apache/drill/common/exceptions/DrillDataReadException.java,
line 26
> > <https://reviews.apache.org/r/32987/diff/2/?file=921681#file921681line26>
> >
> >     I'm inclined not to prefix everything wiht Drill.  Thoughts on this?

I initially went with UserException, DataReadUserException, ParseUserException, UnsupportedOperationUserException...
but SystemUserException doesn't sound right and SystemException sounds like something really
bad did happen


> On April 9, 2015, 6:05 p.m., Jacques Nadeau wrote:
> > common/src/main/java/org/apache/drill/common/exceptions/DrillParseException.java,
line 22
> > <https://reviews.apache.org/r/32987/diff/2/?file=921682#file921682line22>
> >
> >     purpose?  Is this query parsing?  Something else?  If Query Parsing, should
be have line and column?

according to the Exception Types docment: a Parsing Error can include any of the following:
- typos
- missing table
- SQL keyword misuse
- function names/resolution

You can add context using addContext(String) method, for example:
```
parseException.addContext("line 2, column 18");
```
or
```
parseException.addContext("line: 2");
parseException.addContext("column: 18");
```
All context information will be concatenated when generating the user error message


> On April 9, 2015, 6:05 p.m., Jacques Nadeau wrote:
> > common/src/main/java/org/apache/drill/common/exceptions/DrillRemoteException.java,
line 22
> > <https://reviews.apache.org/r/32987/diff/2/?file=921683#file921683line22>
> >
> >     purpose?

If a DrillPBError object is sent passed from a fragment to the Foreman, we used to extract
the error message from this object and then build a new error object, replacing the drillbit
endpoint and creating a new error id along the way.
This class will wrap a DrillPBError object a return instead of building a new error object,
which will preserve the error information sent from the fragments.

This class replaces RemoteRpcException because it was only used for this purpose


> On April 9, 2015, 6:05 p.m., Jacques Nadeau wrote:
> > common/src/main/java/org/apache/drill/common/exceptions/DrillSystemException.java,
line 23
> > <https://reviews.apache.org/r/32987/diff/2/?file=921684#file921684line23>
> >
> >     This should be used only by the internal client layer, right?  If so, let's
make sure the constructors are as protected as possible.

True, we should only create system exceptions through `DrillSystemException.wrapAndSetEndpoint(Throwable,
DrillbitEndpoint)`.
I will make DrillSystemException constructor _private_


> On April 9, 2015, 6:05 p.m., Jacques Nadeau wrote:
> > common/src/main/java/org/apache/drill/common/exceptions/DrillUserException.java,
line 71
> > <https://reviews.apache.org/r/32987/diff/2/?file=921686#file921686line71>
> >
> >     Can we add method like getVerboseMessage() and then choose which one we want
rather than controlling on creation?

Will add it. Still, getMessage() is not the verbose message, it only contains the minimum
amount of information we want to return to the client:
- error type
- error message
- any context information available
- error_id + drillbit_endpoint

getVerboseMessage() will also add the stack trace of the exception that caused the error,
this will be generated when building the DrillPBError object.


> On April 9, 2015, 6:05 p.m., Jacques Nadeau wrote:
> > common/src/main/java/org/apache/drill/common/exceptions/DrillUserException.java,
line 110
> > <https://reviews.apache.org/r/32987/diff/2/?file=921686#file921686line110>
> >
> >     Do we have a separate error object type for bit <> bit?

Do we send error objects between bit <> bit ? DrillPBError is only used between bit
<> Foreman <> client. I am not aware of other error object types.


> On April 9, 2015, 6:05 p.m., Jacques Nadeau wrote:
> > exec/java-exec/src/main/java/org/apache/drill/exec/physical/impl/ScreenCreator.java,
line 181
> > <https://reviews.apache.org/r/32987/diff/2/?file=921689#file921689line181>
> >
> >     Why do we need this?

I will change it to a debug message


> On April 9, 2015, 6:05 p.m., Jacques Nadeau wrote:
> > exec/java-exec/src/main/java/org/apache/drill/exec/physical/impl/broadcastsender/BroadcastSenderRootExec.java,
line 193
> > <https://reviews.apache.org/r/32987/diff/2/?file=921690#file921690line193>
> >
> >     Why do we need this?  If it is a debug message, let's set it at that.

it is a debug message


> On April 9, 2015, 6:05 p.m., Jacques Nadeau wrote:
> > exec/java-exec/src/main/java/org/apache/drill/exec/planner/sql/DrillSqlWorker.java,
line 172
> > <https://reviews.apache.org/r/32987/diff/2/?file=921692#file921692line172>
> >
> >     This seems weird as it looks like I might get  "Failure validating SQL - Sql
Parsing error...".  What will this actually construct?  Can you add to comments here?  Also,
we should probably encapsulate error delineation (" - ") so that we use the same pattern everywhere.

This can definitely be improved, it's just an example of throwing DrillParseException. Using
only the message from the _ValidationException_ we get something like this:

    org.eigenbase.util.EigenbaseContextException: From line 1, column 15 to line 1, column
17: Table 'dfs.data.t.jso' not found

using ex.getCause().getMessage() will give us a much cleaner error message:

    Table 'dfs.data.t.jso' not found

In both cases I think we can safely ignore the "Failure validating SQL", using a DrillParseException
will already display a _PARSE ERROR_ at the beginning of the message


- abdelhakim


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/32987/#review79537
-----------------------------------------------------------


On April 9, 2015, 5:42 p.m., abdelhakim deneche wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/32987/
> -----------------------------------------------------------
> 
> (Updated April 9, 2015, 5:42 p.m.)
> 
> 
> Review request for drill, Jacques Nadeau, Jason Altekruse, and Parth Chandra.
> 
> 
> Bugs: DRILL-2675
>     https://issues.apache.org/jira/browse/DRILL-2675
> 
> 
> Repository: drill-git
> 
> 
> Description
> -------
> 
> **INITIAL PATCH** with a working solution. This patch cleans the path for errors, especially
user errors with meaningful messages, to be propagated properly to the client. 
> 
> The patch includes changes to 2 existing use cases where the error message was successfully
improved.
> 
> The general idea is: if a code wants to throw an exception that contains a meaningful
error message, it throws a DrillUserException. The propagation code will make sure this exception
is propagated to the client. The user exception object doesn't contain the final error message,
but enough information about the error, the client will use this information to display a
better error message.
> Any exception that is not a DrillUserException (or one of it's subclasses) will be considered
as a system exception. For those exceptions the client will only display the error id and
drillbit identity in case the user wants to check the logs for more informations about the
error.
> Error objects sent to the client will still contain a stack trace that can be used to
display more information if the client has enabled the error verbose mode.
> 
> 
> Diffs
> -----
> 
>   common/src/main/java/org/apache/drill/common/exceptions/DrillDataReadException.java
PRE-CREATION 
>   common/src/main/java/org/apache/drill/common/exceptions/DrillParseException.java PRE-CREATION

>   common/src/main/java/org/apache/drill/common/exceptions/DrillRemoteException.java PRE-CREATION

>   common/src/main/java/org/apache/drill/common/exceptions/DrillSystemException.java PRE-CREATION

>   common/src/main/java/org/apache/drill/common/exceptions/DrillUnsupportedOperationException.java
PRE-CREATION 
>   common/src/main/java/org/apache/drill/common/exceptions/DrillUserException.java PRE-CREATION

>   common/src/main/java/org/apache/drill/common/exceptions/ErrorHelper.java PRE-CREATION

>   exec/java-exec/src/main/java/org/apache/drill/exec/ops/FragmentContext.java 18b93e9

>   exec/java-exec/src/main/java/org/apache/drill/exec/physical/impl/ScreenCreator.java
8038527 
>   exec/java-exec/src/main/java/org/apache/drill/exec/physical/impl/broadcastsender/BroadcastSenderRootExec.java
d17fdd4 
>   exec/java-exec/src/main/java/org/apache/drill/exec/physical/impl/partitionsender/StatusHandler.java
5e21878 
>   exec/java-exec/src/main/java/org/apache/drill/exec/planner/sql/DrillSqlWorker.java
b98778d 
>   exec/java-exec/src/main/java/org/apache/drill/exec/rpc/CoordinationQueue.java 0016d6a

>   exec/java-exec/src/main/java/org/apache/drill/exec/rpc/RemoteRpcException.java 14ea873

>   exec/java-exec/src/main/java/org/apache/drill/exec/rpc/RpcBus.java b974963 
>   exec/java-exec/src/main/java/org/apache/drill/exec/rpc/user/QueryResultHandler.java
a1be83b 
>   exec/java-exec/src/main/java/org/apache/drill/exec/store/easy/json/JSONRecordReader.java
cc7cb83 
>   exec/java-exec/src/main/java/org/apache/drill/exec/work/ErrorHelper.java 0773d6c 
>   exec/java-exec/src/main/java/org/apache/drill/exec/work/foreman/Foreman.java 23ef0d3

>   exec/java-exec/src/main/java/org/apache/drill/exec/work/foreman/QueryManager.java 8626d5b

>   exec/java-exec/src/main/java/org/apache/drill/exec/work/fragment/AbstractStatusReporter.java
1b0885d 
>   exec/java-exec/src/main/java/org/apache/drill/exec/work/fragment/FragmentExecutor.java
a7e6c46 
>   exec/java-exec/src/main/java/org/apache/drill/exec/work/fragment/StatusReporter.java
26b5d68 
>   protocol/src/main/java/org/apache/drill/exec/proto/SchemaUserBitShared.java f72d5e1

>   protocol/src/main/java/org/apache/drill/exec/proto/UserBitShared.java 9a9d196 
>   protocol/src/main/java/org/apache/drill/exec/proto/beans/DrillPBError.java ac9cef5

>   protocol/src/main/protobuf/UserBitShared.proto 5e44655 
> 
> Diff: https://reviews.apache.org/r/32987/diff/
> 
> 
> Testing
> -------
> 
> 
> Thanks,
> 
> abdelhakim deneche
> 
>


Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message