Skip to first unread message
Benjamin Geer
Feb 3, 2015, 5:17:19 AM2/3/15
to sta...@clarkparsia.com
Hello,
We are finding that Stardog returns incorrect results when several clients are executing similar SPARQL queries concurrently over HTTP. Here is a minimal test case for reproducing the problem using 10 concurrent client processes:
Steps to reproduce:
1. Make sure that Python 3 and the 'requests' module are installed.
2. Install the provided stardog.properties file in $STARDOG_HOME.
3. Run ./create-test-db.sh to create the test database.
4. Run ./testit.py.
The attached screenshot shows a typical result. It seems that some clients receive a result that incorrectly contains some data intended for another client.
This test case reliably reproduces the problem for us on Mac OS X (single-processor Intel i7) as well as Linux (multi-processor i7).
Best regards,
Benjamin Geer
Digital Humanities Lab
University of Basel
Edgar Rodriguez
Feb 3, 2015, 8:43:50 PM2/3/15
to sta...@clarkparsia.com
Hello Benjamin,
Thank you for the very detailed bug report and tests, I was able to reproduce the behavior described with stardog 2.2.4. The fix will be included in the next release.
Best,
Edgar
--
-- --
You received this message because you are subscribed to the C&P "Stardog" group.
To post to this group, send email to sta...@clarkparsia.com
To unsubscribe from this group, send email to
stardog+u...@clarkparsia.com
For more options, visit this group at
http://groups.google.com/a/clarkparsia.com/group/stardog?hl=en
Benjamin Geer
Feb 4, 2015, 11:20:09 AM2/4/15
to sta...@clarkparsia.com
Dear Edgar,
I'm glad you were able to reproduce the bug. However, since it is very serious, I suggest releasing an immediate patch rather than waiting to fix it in the next release. Not only does it return incorrect data (a critical defect in any DBMS), but the database could become corrupted as a result, if the incorrect data is saved back to the database.
I'm glad you were able to reproduce the bug. However, since it is very serious, I suggest releasing an immediate patch rather than waiting to fix it in the next release. Not only does it return incorrect data (a critical defect in any DBMS), but the database could become corrupted as a result, if the incorrect data is saved back to the database.
Kendall Clark
Feb 4, 2015, 11:28:33 AM2/4/15
to stardog
Benjamin,
Thanks for the feedback.
Cheers,
Kendall Clark
Benjamin Geer
Apr 10, 2015, 8:59:27 AM4/10/15
to sta...@clarkparsia.com
Has this bug been fixed in Stardog 3.0? I didn't see it in the changelog.
Kendall Clark
Apr 10, 2015, 9:10:21 AM4/10/15
to stardog
We reproduced it in 2.2.4 and couldn't reproduce it in 3.0, which is why there's nothing about it in the changelog. Please let us know if yr still seeing it.
Cheers,
Cheers,
Kendall
On Fri, Apr 10, 2015 at 8:59 AM, Benjamin Geer <benjam...@gmail.com> wrote:
Has this bug been fixed in Stardog 3.0? I didn't see it in the changelog.
On Wednesday, 4 February 2015 17:28:33 UTC+1, Kendall wrote:
Benjamin,Thanks for the feedback.Cheers,Kendall Clark
On Wed, Feb 4, 2015 at 11:20 AM, Benjamin Geer <benjam...@gmail.com> wrote:Dear Edgar,
I'm glad you were able to reproduce the bug. However, since it is very serious, I suggest releasing an immediate patch rather than waiting to fix it in the next release. Not only does it return incorrect data (a critical defect in any DBMS), but the database could become corrupted as a result, if the incorrect data is saved back to the database.
Best regards,
Benjamin Geer
Digital Humanities Lab
University of Basel
--
-- --
You received this message because you are subscribed to the C&P "Stardog" group.
To post to this group, send email to sta...@clarkparsia.com
To unsubscribe from this group, send email to
stardog+u...@clarkparsia.com
For more options, visit this group at
http://groups.google.com/a/clarkparsia.com/group/stardog?hl=en
Reply all
Reply to author
Forward
0 new messages