The initial PR for GEODE-8020 is in the support/1.12 branch but has proven to 
reduce performance.  We’ve since found that the buffer corruption that this PR 
was addressing was due to the test in question using TLSv1 as the ssl-protocol. 
 Changing the test to use TLSv1.2 fixed the problem, so I’m planning to revert 
the support/1.12 cherry-pick.

Reply via email to