-
Notifications
You must be signed in to change notification settings - Fork 191
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Adding Missing Test cases and test case refactoring #144
Conversation
# Conflicts: # src/test/java/com/oltpbenchmark/benchmarks/epinions/TestEpinionsLoader.java
i need to do one more pass over these changes at a very low scale factor |
…` is never set. think this is a bug
Please ping me on Slack (otherwise I might miss it). Thanks! |
…reserve ordering; also reducing logging output
… minutes to 6 seconds
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I disagree with your change to NoOp.
@@ -38,7 +38,7 @@ | |||
|
|||
// The query only contains a semi-colon | |||
// That is enough for the DBMS to have to parse it and do something | |||
public final SQLStmt noopStmt = new SQLStmt(";"); | |||
public final SQLStmt noopStmt = new SQLStmt("select 1"); |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This is not the same as a NOOP since it requires evaluation. Is there a reason for this change?
Maybe we should rename it to the "DoNothing" benchmark and then have separate SelectOne
and SemiColon
procedures.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
NoOp was one of the missing test cases; unfortunately HSQLDB doesn't accept ;
as a valid sql statement. my thinking was this was a reasonable fix but now that i'm looking at it i should have just done this with a hsqldb dialect file and not changed the java class. is that a better approach?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
this has been resolved by reverting the change to NoOp.java and instead adding a dialect-hsqldb.xml file with a hsqldb friendly query
|
||
import java.util.List; | ||
|
||
@Ignore("the testcase is under development") |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Is this still true?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
as soon as i can submit my config-refactor pr, this is greatly simplified. I marked this as ignore because at the moment it only loads the schema for one of the benchmarks instead of both. fixing this was a lot of the work that i'd already done in config-refactor. my thinking is to ignore this test until the subsequent pr is merged and then it can be reenabled.
src/test/java/com/oltpbenchmark/benchmarks/chbenchmark/TestCHBenCHmarkWorker.java
Show resolved
Hide resolved
src/test/java/com/oltpbenchmark/benchmarks/twitter/TestTwitterWorker.java
Show resolved
Hide resolved
…e but possible port conflicts; also waiting during stop incase there is a problem brining hsqldb down
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM. Please resolve conflicts
The test package was missing a number of benchmark tests. This PR attempts to resolve all of that by adding any missing benchmark, loader or worker tests. To simplify the creation of these tests, some refactoring was done to the base "abstract" test classes. 2 tests, chbenchmark and twitter, have been temporarily
@Ignore
d as the changes required to make them work have already been addressed in PR #134 which.Of course when adding the missing test cases, some issues were identified in either the benchmarks themselves (usually having to do with behavior at very low SF, or with the HSQLDB dialect files). These issues were resolved as part of this PR.
Furthermore, I changed the way HSQLDB is initialized so that tests can be paused and the database can be accessed/queried. This is extremely useful for troubleshooting.