#opendaylight-meeting: carbon release sync

Meeting summary

1. recap (colindixon, 15:06:52)
   1. the biggest issue with Carbon was that Karaf 4 seemed to have wildly changed bundle/feature lifecycle managment from Karaf 3 and that would cause all kinds of problems (colindixon, 15:07:26)
   2. in particular, it reloads bundles and features in ways that Karaf 3 didn't, many (really, almost all) ODL bundles and features dont' handle being reloaded correctly, so pretty regularly a bundle will not come back and result in things just hanging (colindixon, 15:08:34)
   3. this happens with Karaf 4 SingelFeatureTests pretty regularly causing the autorelease to fail almost every time on those (colindixon, 15:08:52)
   4. this has more dire implications when doing normal things with CSIT and installing feautre (colindixon, 15:09:09)
   5. https://lists.opendaylight.org/pipermail/release/2017-May/010746.html the result is that combined with a statement from Jamie that Karaf 3 will have security support for the duration of Carbon, people are thinking of shipping Carbon on Karaf 3 and moving to Karaf 4 as a major effort in Nitrogen (colindixon, 15:10:15)
   6. it looks like there are 7 +1s from TSC members to move Carbon to Karaf 3 (colindixon, 15:10:37)
   7. https://docs.google.com/spreadsheets/d/1VcB12FBiFV4GAEHZSspHBNxKI_9XugJp-6Qbbw20Omk/edit#gid=1455372448 colindixon has been tracking autorelease failures here (colindixon, 15:21:14)
   
   
2. where we are (colindixon, 15:30:07)
   1. we have 3 autorelease jobs for carbon now: normal, failnever, and notest (colindixon, 15:30:31)
   2. ACTION: zxiiro-away will work on consolidating those into one job that runs with -Dskip.karaf.featureTest=true for now and maybe a better way to disable Karaf 4 in the future (colindixon, 15:44:57)
   3. ACTION: colindixon will close the thread with 7 +1s from TSC members on shipping Carbon on Karaf 3 asking for any critical feedback on the decisions (colindixon, 15:45:24)
   4. ACTION: colindixon will follow up with a mail to relaese and TSC basically saying we should take the last succsessful staging repository and use it as RC0 and lock the stable/carbon branches to start making progress on testing and releasing Carbon (colindixon, 15:46:17)

Action items

1. zxiiro-away will work on consolidating those into one job that runs with -Dskip.karaf.featureTest=true for now and maybe a better way to disable Karaf 4 in the future
2. colindixon will close the thread with 7 +1s from TSC members on shipping Carbon on Karaf 3 asking for any critical feedback on the decisions
3. colindixon will follow up with a mail to relaese and TSC basically saying we should take the last succsessful staging repository and use it as RC0 and lock the stable/carbon branches to start making progress on testing and releasing Carbon

Action items, by person

1. colindixon
   1. colindixon will close the thread with 7 +1s from TSC members on shipping Carbon on Karaf 3 asking for any critical feedback on the decisions
   2. colindixon will follow up with a mail to relaese and TSC basically saying we should take the last succsessful staging repository and use it as RC0 and lock the stable/carbon branches to start making progress on testing and releasing Carbon

People present (lines said)

1. colindixon (15)
2. odl_meetbot (3)