Hi, To conclude, after doing a lot of work in debugging we were able to reduce the deployment of our VRs from ~2 hours to ~5 MINUTES.
Two PRs are open for this against the 4.9 branch: - https://github.com/apache/cloudstack/pull/2077 - https://github.com/apache/cloudstack/pull/2089 The problem was that in Basic Networking each VR would get ALL DHCP information instead of just the information for that POD (PR 2077). The other issue is that for each entry dnsmasq and Apache would be restarted combined with some other things. By delaying this to the end of the provision of the router we save a lot of time. Both PRs are running in production on our cloud in Basic Networking with a few thousands Instances behind it. No problems found so far. Wido > Op 2 mei 2017 om 19:57 schreef Wido den Hollander <w...@widodh.nl>: > > > Hi, > > Last night I upgraded a CloudStack 4.5.2 setup to 4.9.2.0. All went well, but > the VR provisioning is terribly slow which causes all kinds of problems. > > The vr_cfg.sh and update_config.py scripts start to run. Restart dnsmasq, add > metadata, etc. > > But for just 1800 hosts this can take up to 2 hours and that causes timeouts > in the management server and other problems. > > 2 hours is just very, very slow. So I am starting to wonder if something is > wrong here. > > Did anybody else see this? > > Running Basic Networking with CloudStack 4.9.2.0 > > Wido
