Bruce, Here is what I am seeing in my test (same one I sent you).
904 server -> 823 client: Before writing to stream: Client Message Dispatcher for 192.168.2.4(client:78763:loner):49721:b274160b:client(version:GFE 8.1): ClientUpdateMessageImpl.getGFE70Message membershipIdLength=39; membershipId=[4, -64, -88, 2, 4, 0, 0, 4, 3, 5, 10, 87, 0, 1, 57, 87, 0, 4, 112, 101, 101, 114, -80, 114, 25, 102, 117, -114, -43, -72, -98, -85, -124, -126, 97, -21, -116, 115, 0] Being written to stream: Client Message Dispatcher for 192.168.2.4(client:78763:loner):49721:b274160b:client(version:GFE 8.1): EventID.toData writing membershipIdLength=39; membershipId=[4, -64, -88, 2, 4, 0, 0, 4, 3, 5, 10, 87, 0, 1, 57, 87, 0, 4, 112, 101, 101, 114, -80, 114, 25, 102, 117, -114, -43, -72, -98, -85, -124, -126, 97, -21, -116, 115, 0] 91 server -> 823 client: Before writing to stream: Client Message Dispatcher for 192.168.2.4(client:78763:loner):49721:b274160b:client(version:GFE 8.1): ClientUpdateMessageImpl.getGFE70Message membershipIdLength=39; membershipId=[4, -64, -88, 2, 4, 0, 0, 4, 3, 5, 10, 87, 0, 1, 57, 87, 0, 4, 112, 101, 101, 114, -80, 114, 25, 102, 117, -114, -43, -72, -98, -85, -124, -126, 97, -21, -116, 115, 0] Being written to stream: Client Message Dispatcher for 192.168.2.4(client:78763:loner):49721:b274160b:client(version:GFE 8.1): EventID.toData writing membershipIdLength=22; membershipId=[4, -64, -88, 2, 4, 0, 0, 4, 3, 5, 10, 87, 0, 1, 57, 87, 0, 4, 112, 101, 101, 114] In the 904 case, the bytes in the original membership id are written to the stream. In the 91 case, the last 17 bytes are removed before the membership id is written to the stream. Thanks, Barry Oglesby On Mon, Jul 3, 2017 at 1:14 PM, Bruce Schuchardt <bschucha...@pivotal.io> wrote: > I've checked in the fix for geode-3153, though Barry said he may still be > seeing trouble with it. I plan to get together with him on-line wednesday > morning, if possible, to step through his test & verify if the problem > still exists. > > I've also pushed the commit to release/1.2.0 > > > > On 7/1/17 8:54 AM, Bruce Schuchardt wrote: > >> This is a 1.2.0 blocker. It's passed precheckin and new >> backward-compatibility unit tests and the reviewboard code review has no >> outstanding issues. I'd like to get a shipIt or two & I'll push the >> changes to develop and release/1.2.0. >> >> https://reviews.apache.org/r/60570/ >> >> >
