> On Dec. 22, 2015, 1:55 a.m., Puneet Gupta wrote:
> > lens-server/src/main/java/org/apache/lens/server/query/LensServerDAO.java, 
> > line 162
> > <https://reviews.apache.org/r/41506/diff/1/?file=1173734#file1173734line162>
> >
> >     This can be a static string. We can create it once and reuse. Same can 
> > be follwed for other instances too.

Dropping . This string is already Internalized.


> On Dec. 22, 2015, 1:55 a.m., Puneet Gupta wrote:
> > lens-server/src/main/java/org/apache/lens/server/query/LensServerDAO.java, 
> > line 207
> > <https://reviews.apache.org/r/41506/diff/1/?file=1173734#file1173734line207>
> >
> >     driverendtime can be skipped from ordering construct.
> 
> Rajat Khandelwal wrote:
>     Added it for a currently impossible (but plausible) bug, when driver 
> start time somehow gets missed while updating in the table, but driver end 
> time is updated correctly. Keeping both in order by will allow the attempts 
> to be ordered. Let me know if you still feel it should be removed.

Should we have a separate column then , something like attempt_num. 
It ll be more clear when we look at the DB Table.


> On Dec. 22, 2015, 1:55 a.m., Puneet Gupta wrote:
> > lens-server/src/main/java/org/apache/lens/server/query/QueryExecutionServiceImpl.java,
> >  line 834
> > <https://reviews.apache.org/r/41506/diff/1/?file=1173735#file1173735line834>
> >
> >     Shall we let the driver decide on number of re-tries ? It can be diff 
> > for Hive and JDBC. Eventually (in future) it can even differ based on error 
> > type.
> 
> Rajat Khandelwal wrote:
>     Sure, but I believe the server should impose some upper bound on retries. 
> That's why I've kept default retries to 3. That's server's bound, the drivers 
> have full freedom apart from that.

Not sure if server should restrict the driver. A driver may decide for a 
particualr query/error it can try more than 3 times. The driver's shouldRetry() 
method can decide in that case ?


> On Dec. 22, 2015, 1:55 a.m., Puneet Gupta wrote:
> > lens-server/src/main/java/org/apache/lens/server/query/QueryExecutionServiceImpl.java,
> >  line 1857
> > <https://reviews.apache.org/r/41506/diff/1/?file=1173735#file1173735line1857>
> >
> >     Should we store numOfAttempts in finishedQueries table?  If the value 
> > is zero, we need not check failed attempts table at all. Also admin can 
> > query finished query table and gets attempts statistics easily.  
> >     
> >     Or is there a way to add a new Query in lensServerDao which joins 
> > finished queries table with failed attempts and gets the result in one hop. 
> > It can be messy though.
> 
> Rajat Khandelwal wrote:
>     I don't think it'll be messy. `select * from finished_queries where 
> handle='handle'` will now be written as `select *, (count(*) from 
> failed_attempts where handle='handle') failed_attempts from finished_queries 
> where handle='handle'`
>     
>     Though it *will* be messy without the handle filter which it will still 
> be even after addition of the `num_attempts` field.

I meant that can we can query the DB just once for 99% cases where attempts are 
0 ? As of now we have to query twice in all cases. Once finished queries table 
and once fialed attempts table.


- Puneet


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/41506/#review111459
-----------------------------------------------------------


On Dec. 21, 2015, 9:15 a.m., Rajat Khandelwal wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/41506/
> -----------------------------------------------------------
> 
> (Updated Dec. 21, 2015, 9:15 a.m.)
> 
> 
> Review request for lens.
> 
> 
> Bugs: LENS-899
>     https://issues.apache.org/jira/browse/LENS-899
> 
> 
> Repository: lens
> 
> 
> Description
> -------
> 
> 
> Diffs
> -----
> 
>   lens-api/src/main/java/org/apache/lens/api/query/QueryStatus.java 
> 91cbe391a2bc81edae587b268104940122993243 
>   
> lens-server-api/src/main/java/org/apache/lens/server/api/LensConfConstants.java
>  88e5a01e3f08b12ff9e872d7412496dfc70ea626 
>   
> lens-server-api/src/main/java/org/apache/lens/server/api/driver/AbstractLensDriver.java
>  ed1fc43ab6e6ef8c34c5c2510cd5f5333ca16179 
>   
> lens-server-api/src/main/java/org/apache/lens/server/api/driver/DriverQueryStatus.java
>  79d8bf1e8f9798714d14e5a0336e24a148c4ca27 
>   
> lens-server-api/src/main/java/org/apache/lens/server/api/driver/LensDriver.java
>  3d38ddd2dfe596b348bb48807aac5caeab5e936a 
>   
> lens-server-api/src/main/java/org/apache/lens/server/api/query/FailedAttempt.java
>  PRE-CREATION 
>   
> lens-server-api/src/main/java/org/apache/lens/server/api/query/FinishedLensQuery.java
>  a57a6e4ed1827ee5094cbfd423c183766d9dc5f4 
>   
> lens-server-api/src/main/java/org/apache/lens/server/api/query/QueryContext.java
>  1269e45ff7e5b9392108e338735b9774fb8a2686 
>   lens-server/src/main/java/org/apache/lens/server/query/LensServerDAO.java 
> d8e654d72989d7a8f9334adc81b82a9e0810ee39 
>   
> lens-server/src/main/java/org/apache/lens/server/query/QueryExecutionServiceImpl.java
>  2dff9afeb384738fa21ef7381cb5a54e3c02165a 
> 
> Diff: https://reviews.apache.org/r/41506/diff/
> 
> 
> Testing
> -------
> 
> 
> Thanks,
> 
> Rajat Khandelwal
> 
>

Reply via email to