==================================== #opnfv-meeting: Octopus Team meeting ==================================== Meeting started by ulik at 14:07:05 UTC. The full logs are available at http://ircbot.wl.linuxfoundation.org/meetings/opnfv-meeting/2015/opnfv-meeting.2015-03-23-14.07.log.html . Meeting summary --------------- * yes we can start with Recap on test & Performance meeting (rprakash, 14:08:06) * - Recap from Test&Performance meeting (ulik, 14:11:57) * https://etherpad.opnfv.org/p/octopus (fdegir, 14:12:33) * Look for How to Connect Servers from Labs to OPNFV Jenkins (fdegir, 14:12:47) * Fatih explains that the current plan opens the LF lab so jenkins slaves can listen to the central master (ulik, 14:12:50) * Short term solution -Labs opening firewall on Server using JNMP - its on opnfv site (any links?) (rprakash, 14:13:20) * there may be some limitations with some permission issues (ulik, 14:13:22) * The method explained in Etherpad is called JNLP (fdegir, 14:13:50) * and a server from Ericsson lab is currently connected to OPNFV Jenkins (fdegir, 14:14:08) * https://build.opnfv.org/ci/ (fdegir, 14:14:41) * Fatih shows how Ericsson lab connects to LF jenkins (ulik, 14:15:28) * This way, a lab has full control, e.g. disconnect for maintenance purpose (ulik, 14:16:54) * Fatih encourages to test this by connecting any local jenkins slave. (ulik, 14:17:39) * Intel lab is using this (ulik, 14:18:06) * no action open on octopus side now. Test and Performance team needs to come back to us (ulik, 14:20:28) * Question about labs talking to each other. Fatih doesn't expect this at the moment. (ulik, 14:21:15) * Iben explains that for ODL, Spirent listens to the artifacts repository. (ulik, 14:21:54) * Fatih thinks this might be good for later releases but not release 1. (ulik, 14:22:18) * Define way to use hartdware (ulik, 14:25:14) * Jerry explains which jobs would have hardware dependencies or which would better run in virtual environment (ulik, 14:26:58) * Fatih explains we have the necessary information only from Fuel track. (ulik, 14:27:23) * Jerry explains from experience using virtual machines for the builds and the way to use VLANs (ulik, 14:34:12) * LINK: https://wiki.opnfv.org/get_started/foreman_quickstack_status (rprakash, 14:36:42) * Tim Rozet reports from BGS how they use network during deploy (ulik, 14:37:15) * is thre a topology digram for the configuration you expalined now? (rprakash, 14:40:15) * one deploy server could serve multiple PODs (ulik, 14:41:50) * see Topology section: https://wiki.opnfv.org/get_started_experiment1 (trozet, 14:42:14) * there will be no issue when a lab uses same installer. The issue comes up, when using different installers in LF environment. (ulik, 14:43:28) * Iben explains from experience at Spirent for this. (ulik, 14:45:21) * Iben will share some documentation with the team. (ulik, 14:46:26) * Iben sure we like to see that on opnfv site with links added for different labs (rprakash, 14:46:48) * artifact repository (ulik, 14:48:27) * Google Storage to be used for OPNFV Artifacts (rprakash, 14:49:43) * https://wiki.opnfv.org/wiki/artifactrepos#what_will_be_used_as_artifact_repository_for_opnfv (fdegir, 14:50:13) * aric explains the possibilities with Google store (ulik, 14:50:45) * We will need to be able to pick the latest artifact from Google Storage (fdegir, 14:51:36) * This will be made available while we move (fdegir, 14:51:55) * Prakash asks why not use Swift on LF hardware and Aric explains that we don't have big hard drives (ulik, 14:52:33) * Google Storage is blocked in China - this may require further thinking (fdegir, 14:53:11) * Jerry explains that Google is not available from China (ulik, 14:53:20) * Another shortage are missing backups (ulik, 14:53:41) * Google shares between different locations, but there is no SLA. (ulik, 14:54:41) * this could be sufficient, since build artifacts can be rebuilt if necessary. (ulik, 14:55:02) * as long as we have full traceablity (fdegir, 14:55:17) * LINK: http://storage.googleapis.com/opnfv-test/opnfv-P0000.iso (aricg, 14:56:23) * LINK: http://storage.googleapis.com/opnfv-test/opnfv-P0000.iso (rprakash, 14:58:11) * mirror this repo on each lab if acceptable to all (rprakash, 14:58:52) * ACTION: Jerry, Iben, Fatih, to think about how to setup mirrors or other to solve this. (ulik, 14:59:29) Meeting ended at 15:00:21 UTC. Action items, by person ----------------------- * iben * Jerry, Iben, Fatih, to think about how to setup mirrors or other to solve this. People present (lines said) --------------------------- * ulik (33) * fdegir (12) * rprakash (8) * collabot (5) * alexpilotti (3) * aricg (2) * trozet (2) * iben (1) * dfarrell07 (1) * jerryz (0) Generated by `MeetBot`_ 0.1.4