* Allow any recursion and cache queries for named svc
* Bump maas v3 to the actual version
Signed-off-by: Ruslan Aliev <raliev@mirantis.com>
Change-Id: I16a4ec843dc73a2349e8603d4200920599eab918
Provide a knob to adjust some less-common MAAS configuration settings.
Changes the default values as follows: disables network discovery, sets
the active subnet mapping interval to 0 (from 10800 seconds), marks the
intro as completed, and disables Google analytics.
Refer to `maas $PROFILE maas set-config -h` for the list of available
configuration items.
Change-Id: I46d348ef5777e22ebeb7a062e5f6061d9ad61a1c
- The import script would skip creating a new boot source
selection for a non-default distro in some cases due
to a non-recommended if construct. Change to the recommended
'if ! grep -q' pattern
Change-Id: I59e6732598f74fc34a6986dbdfe4200d8cd9ea9f
- The import resources job waits for rack controllers to sync
the imported resources. In an environment that rack pods have
been scheduled away from a node, it leaves a dead rack controller
in the MAAS database. We cannot gate on dead controllers syncing
as rackd is no longer running, so now only look at rack controllers
with a running rackd.
Change-Id: I5ca16a0c97ed201a08844ca7c82c2cbb7d059aa7
Retry logic was added in the check_then_set function to avoid edge case
scenarios where the rack controller is not synced yet.
Change-Id: I33fe7aabf32b872d67c31616ee6074d262aece4a
- The import resources job can sometimes take much longer
than expected seemingly because even though the rack
controller has the image data, it doesn't report it.
Rather than waiting for MAAS to resync the rack controller
automatically, explicitly start an image sync on each rack
controller that isn't synced
Change-Id: Ie3a8124db2266380420b1efa4aaec6e3aa3c2904
- If the import job triggers before the rack controller pod(s)
have registered w/ the region controller, then it must fully timeout
and then reschedule to pass. Update it so that each time it checks for
rack controller image sync, it updates the list of all registered
rack controllers.
- Update register service to be part of Dockerfile so it can be
enabled.
Change-Id: I72e190d472ad259da65b2e583b2a16d8adf660f5
Looks like new version of MAAS has fixed long standing bug
https://bugs.launchpad.net/maas/+bug/1779712
This will match internal MAAS ports to NodePort.
Change-Id: I639a4c492eb80545c69fd132d3b2dc4cca524933
- Add a selection to explicitly choose the Ubuntu release
- Add a gate for rack controller image sync before attempting to
configure ephemeral image selection
Change-Id: Id8397d79fa5d136d78923f838c624283fad3d769
- Add an improved helm test for MaaS to validate that
there are available images for bootstrapping and that
at least one rack controller is available
- Improve idempotence of the import images job for more
resilient chart upgrades
Change-Id: Id1b07981f13636fcd3b5b1d0bf5fb2ce0f3e5216
- Relied on JSON field names being in a deterministic order. This
is not guaranteed. Update import job to no longer assume this.
Change-Id: I8d0a3abae94b99a94df61346b504517e2cb4a8ca
- Fix bash conditional error causing the export
key job to fail for any chart upgrade
- Enhance the image import/configuration job to
be more robust with additional retrying and better
failure detection
Change-Id: I874a98587a56b03905d740cd53cdd70a8419a04e
- Allow overrides to configure the default operating
system, distribution and kernel to use in MAAS
- The defaults will be used for node enlistment and
commissioning and for deployment if per node overrides
are not used
Change-Id: I872df1c7b7741ec5e599a7a1d05c2707afd6d99e
- If conf.cache.enabled is true, deploy a sidecar container
in the region pod with a simplestreams repo populated w/ a Ubuntu image
- If conf.cache.enabled is true, configure MaaS to source the image
from the sidecar
- Update README
Closes #1
Change-Id: I968614d6fb7ca86589dc6e2efd1f66ae920d03a8