I'm not sure what to make of this discussion. Han, will you update the patch to update the documentation? It seems that, at a minimum, that is the conclusion here. On Wed, Aug 14, 2019 at 06:40:45PM -0700, aginwala wrote: > Sure. Thanks for re-validating different corner cases with me. Yup, we
Sure. Thanks for re-validating different corner cases with me. Yup, we can update more details in leader-only section about using single LB VIP while accessing clustered db via ovn-nbctl/ovn-sbctl for sure to avoid confusion. On Wed, Aug 14, 2019 at 3:21 PM Han Zhou wrote: > Hi Aginwala,
Hi Aginwala, thanks for the testing. The change of this patch will cause the IDL to avoid connecting to a follower if "leader_only" is set to "true". It is the same behavior as before when multiple remotes are used, but now it just does the same even when a single remote is used, because the
On Tue, Aug 13, 2019 at 7:07 PM aginwala wrote: > Thanks for the fixes. Found a bug in current code which breaks both nbctl > with local socket and daemon mode on follower nodes. Also nbctl daemon mode > always tries to connect to leader node by default which also failed to > connect. > e.g.
From: Han Zhou When clustered mode is used, the client needs to retry connecting to new servers when certain failures happen. Today it is allowed to retry new connection only if multiple remotes are used, which prevents using LB VIP with clustered nodes. This patch makes sure the retry logic