Prevent long overflow in SNI address resolution #2035
Conversation
| @@ -64,7 +64,7 @@ public InetSocketAddress resolve() { | |||
| // The order of the returned address is unspecified. Sort by IP to make sure we get a true | |||
| // round-robin | |||
| Arrays.sort(aRecords, IP_COMPARATOR); | |||
| int index = (aRecords.length == 1) ? 0 : (int) OFFSET.getAndIncrement() % aRecords.length; | |||
| int index = (aRecords.length == 1) ? 0 : Math.abs((int) OFFSET.getAndIncrement()) % aRecords.length; | |||
There was a problem hiding this comment.
Why is OFFSET an AtomicLong, shouldn't it be an AtomicInteger?
There was a problem hiding this comment.
Good catch, changing.
lukasz-antoniak
force-pushed
the
sni-overflow
branch
from
8a57448 to
d68ec67
Compare
core/src/main/java/com/datastax/oss/driver/internal/core/metadata/SniEndPoint.java
Outdated
Show resolved
Hide resolved
|
Jenkins build failed due to formatting issues. Just run "mvn com.coveo:fmt-maven-plugin:format" and commit the results... that should be good enough. |
lukasz-antoniak
force-pushed
the
sni-overflow
branch
from
9594562 to
4cdfa39
Compare
|
Done. |
|
Kicked off Jenkins tests again now that formatting has been resolved |
|
I keep wanting to try to make this work with IntStream but the concurrency win from using AtomicInteger is just too big a benefit to overcome. I'll wait until the Jenkins run finishes but assuming there's nothing unexpected there I'm good with this. |
|
Re-run of Jenkins build after format fixes was as green as expected. There were a few test failures but they were all of the CASSJAVA-95 variety. @lukasz-antoniak can you update the commit message for this guy to include the "patch by" line? I think once we have that this is ready to go in. |
patch by Lukasz Antoniak and Alexandre Dutra; reviewed by Bret McGuire
lukasz-antoniak
force-pushed
the
sni-overflow
branch
from
4cdfa39 to
965f0dd
Compare
Prevent overflowing to negative value.