This allows the system to converge during corner cases where there is an
error, or when there are no remotes being used, but we are using the
--no-watch variable.
I deliberately left this in as a separate commit instead of rebasing
into the remote execution development branch because the placement of
the Unregister() and semaphore.V(1) were quite subtle and easy to forget
about.
This patch extends the --converged-timeout argument so that when used
with --remote it waits for the entire set of remote mgmt agents to
converge simultaneously before exiting.
purpleidea says: This particular part of the patch probably took as much
work as all of the work required for the initial remote patches alone!
This implements the new extensions to the converged UUID API so that we
can keep a consistent timer running and reset it when needed. This is
useful because we now allow certain Watcher callbacks and other
operations to explicitly _not_ cause a convergerUUID to un-converge.
This is a necessary dependency for the distributed remote
converged-timeout work.
This adds a new method of marking whether a particular UUID has
converged or not. You can now Start, Stop, or Reset a convergence timer
on the individual UUID's. This wraps the existing SetConverged calls
with a hidden go routine. It is not recommended to use the SetConverged
calls and the Timer calls on the same UUID.
This adds new helper methods to the API such as the ability to query the
set timeout, and to set the state change function after initialization.
The first makes it easier to pass in the timeout value to functions and
structs, because you don't need to pass it separately from the converger
object. The second makes it easy to specify the state change functon
when you don't know what it is at creation time. In addition, it is more
powerful now, and tells you when we converge or un-converge in case we
want to take different actions on each.
We updated the API and behaviour to make two changes:
1) We remove the log.* stuff, and replace the Fatal calls with straight
calls to panic(). These are meant to be like an assert, and shouldn't
happen unless there is a user error or a bug.
2) We made the !uuid.IsValid() checks return an error instead of causing
a panic. These returns errors instead, and makes the process safer for
users who are okay calling a function that won't have an effect.
If you don't give your two host cluster enough time to "feel healthy",
it will generate an error if you do operations within five seconds. This
is a regression and the five seconds is also quite arbitrary. This is
detailed at: https://github.com/coreos/etcd/issues/6305
This seems to be a bit of a race condition, even with a 10s timer, so
this also disables the StrictReconfigCheck. Re-enable this as soon as
possible.
This is a new mode to be used for bootstrapping mgmt clusters or in
situations with tight operational restrictions.
This includes the basics, additional functionality will follow!
If the $(shell ...) command fails, the ':=' operator fails as well
preventing variable overrides from functioning.
Wrap the assignment in an $(or ...) to prevent the shell script from
running if the variable is set from the command line or the environment.
Fixes issue #58
According to the documentation for uname:
-m, --machine
print the machine hardware name
-i, --hardware-platform
print the hardware platform (non-portable)
So use the portable -m version.
The MemberList() function which apparently looks at all of the endpoints
was blocking right after a member exited, because we still had a stale
list of endpoints, and while a member that exited safely would update
the endpoints lists, the other member would (unavoidably) race to get
that message, and so it might call the MemberList() with the now stale
endpoint list. This way we invalidate an endpoint we know to be gone
immediately.
This also adds a simple test case to catch this scenario.
