FULL PRODUCT VERSION :
java version "1.7.0_79"
OpenJDK Runtime Environment (IcedTea 2.5.5) (7u79-2.5.5-0ubuntu0.14.04.2)
OpenJDK 64-Bit Server VM (build 24.79-b02, mixed mode)

FULL OS VERSION :
Linux prod-catalyst-all-01 3.13.0-49-generic #83-Ubuntu SMP Fri Apr 10 20:11:33 UTC 2015 x86_64 x86_64 x86_64 GNU/Linux


A DESCRIPTION OF THE PROBLEM :
After many weeks of operating normally, today on multiple hosts which don't communicate with one another, we experienced many VMs suddenly and rapidly increasing in their virtual memory consumption from around 4GB to more than 35GB, while the Java heap size was still just ~2GB.

It isn't clear what might have instigated this unexpected memory growth, but once it exceeded about 32GB, it began to cause other processes on the hosts to fail due to lack of resources.

Pmaps and smaps were gathered and show a large number (32000+) of 1016kB allocations.

THE PROBLEM WAS REPRODUCIBLE WITH -Xint FLAG: Did not try

STEPS TO FOLLOW TO REPRODUCE THE PROBLEM :
It isn't clear what caused the problem to manifest today, but it happened on the same day on multiple hosts.

EXPECTED VERSUS ACTUAL BEHAVIOR :
Actual: Virtual process size grew massively beyond the norm unexpectedly.

Expected: Virtual process size might grow some over time, but shouldn't spike and exhaust all resources on the machine.
ERROR MESSAGES/STACK TRACES THAT OCCUR :
Link to pmap: https://bluejeans.box.com/s/r67ejhrnvucc74fvi14uoz867k6spuzt
Link to smap: https://bluejeans.box.com/s/82cio0lo4sbckklya9zcyr9gg3cpfz3p

REPRODUCIBILITY :
This bug can be reproduced rarely.