Commits on Jul 17, 2020

1. Improve bootstrap reliability on heterogeneous UPI network configurat… …

   …ions
   
   Before this change, bootstrap IP discovery assumed that the first address of the
   unicast interface must be the bootstrap IP. This assumption doesn't always hold
   in the face of user-defined interfaces and addresses whose ordering isn't
   guaranteed. When the assumptions are broken and the incorrect bootstrap IP is
   selected, bootstrapping fails because quorum cannot be established.
   
   This change improves the accuracy of bootstrap IP discovery by more flexibly
   accounting for a wider variety of possible network interface configurations.
   
   An IP is now considered the bootstrap IP if all of the following are true.
   
   For IPv4:
   
   * The IP is contained by the machine CIDR defined in the cluster configuration
   * On bare metal platforms, the IP is not the API or DNS VIP in the cluster configuration
   
   For IPv6, the same must be true in addition to the following:
   
   * The IP is not deprecated
   * The IP is routable according at least one non-default route
   
   This work is adapted from https://github.com/openshift/baremetal-runtimecfg/blob/master/pkg/utils/utils.go.

   

   ironcladlou committed Jul 17, 2020
   Configuration menu

   • View commit details
   • Copy full SHA for 002e22a
   • Browse repository at this point

   Copy the full SHA

   002e22a View commit details

   Browse the repository at this point in the history

2. Make bootstrap IP discovery backwards compatible with previous assump… …

   …tions
   
   Before this patch, the new bootstrap IP discovery mechanism would fail bootstrapping
   if no IP could be intelligently discovered. A side-effect of that is effectively
   validating the machine network CIDR by asserting the bootstrap IPs ability to be
   discovered within it. Because there may still be edge cases where we fail to detect
   but where the old assumption to choose the "first IP" would still work, we could
   introduce an undue burden to fix all existing uses of machine network CIDR even
   when our fallback could continue to work in those cases.
   
   This patch adds a fallback behavior so that when intelligent discovery fails, the
   first listed IP is selected with a warning, preserving the original discovery
   behavior.
   
   This does effectively mean that clusters can still fail to bootstrap if even the
   first IP assumption is wrong, but we can presumably use those failures to further
   improve detection.
   
   A worthwhile future improvement would be to find a way to more loudly and clearly
   surface to the user when we're blindly guessing about the IP, as the resulting
   downstream failure may obfuscate the source of failure if bootkube logs are lost.

   

   ironcladlou committed Jul 17, 2020
   Configuration menu

   • View commit details
   • Copy full SHA for 68ef6a9
   • Browse repository at this point

   Copy the full SHA

   68ef6a9 View commit details

   Browse the repository at this point in the history