1 of 1 people found this helpful
I appologize, that's my fault. I've been in the process of moving our internally created tool to the opensource and have not finished. We've been using this tool since the MetaMatrix days and have been trying to push this out. It would do what you're asking for. If you're interested and if you'll give me a few day's, I'll push it up in priority and get it done.
Ok, the latest changes have been applied and updated the README.md to help explain how to use it.
thank you for the quick reply.
It's good to know that this tool has the support of the community. I am very interested in making use of the tool and I am more than happy to contribute, if not in the coding then at least in the documentation.
Then I'll finish up on the porting process and then I'll get back to you when its ready. And then we take it from there.
Ok, the latest changes have been applied to the git repo and an updated README.md to help explain how to use it.
thank you for getting back with the changes so quickly.
I am still having some trouble making the tool work, and I would like to confirm a few things that I noticed:
- I had to install the latest Java SDK (was using 1.6x) to complete the Maven build successfully. Is this a known dependency?
- I had to change the classpath to point to the generated class files, otherwise the script is complaining that it couldn't find the main class org.jboss.bqt.client.TestClient. (I commented out the existing line and replaced it with a new one)
- I am running the tool locally, on a Windows machine, using Cygwin. The dev machine, where the Teiid instance is running, is a remote Unix box. I have configured the "run.conf", as well as the .properties files, to point to the right location, but I am getting the following error:
- One of the steps in the README file says "place your jdbc driver into the lib directory". There is no /lib folder in the BQT, however I did specify the location of the driver in the CP. It maked me wonder though whether I should be running this tool on the actual machine where the VDB is deployed.
- for JDK, that dependency can be removed from the pom
- ah, it appears you're trying to run the tests thru maven. The initial release only supports using the distro (distro/target/distribution). The enhancements can easily be made to do use maven.
- the error your getting is because the scenarios directory is not found. Try using the distro, its packaged in the kit, as well as, the lib directory.
atzakas: to build the distro: mvn clean install -Pdistro
thanks for the help. I can now run my queries in bulk and do some regression testing.
Just a couple of suggestions:
a) With regards to the capability of capturing the query execution time, is it possible to get a more detailed view of how long it took to access each federated data source?
Knowing how long it takes to execute the entire query is nice, but it would be more useful if I could tell how long it took for each translated/decomposed query to execute.
b) Would it be easy to add an option to generate a log file containing the explain query plan for each of the queries included in the tests? Along with the detailed timings from a) this would make a powerful tool for troubleshooting and optimizing queries.
a) The query plan does contain information about the node processing time, I'll need to check to see how to access it in the PlanNode. But maybe doable.
b) this could be a good feature to add. Have an option that will write out the query plan, and then use it in regression testing. Because if you do any specific tuning on a query (i.e., add hint), it would be helpful to know if the query plan ever changes.
I'll log an enhancement for "b", and I'll do a followup to see what's possible on "a".