After some investigation, it appears that SocketSet.Shutdown() and
SocketSet.Close() are not synchronous operations. The sendto system call
called in SocketSet.Shutdown() is not a blocking send. That means there
is a race in which SocketSet.Shutdown() sends a message to a file
descriptor to unblock select, while SocketSet.Close() will close the
file descriptor that the message is being sent to. If SocketSet.Close()
wins the race, select is listening on a dead file descriptor and will
hang indefinitely.
This is fixed in the current master by putting SocketSet.Close() inside
of the goroutine in which data from the socket is being received. It
relies on SocketSet.Shutdown() being called to terminate the goroutine.
While this works most of the time, there is a race here. All the
goroutines can also be terminated by a closeChan. If the goroutine
receives an event (thus unblocking select) and then closeChan is
triggered, both SocketSet.Shutdown() and SocketSet.Close() race, leading
to undefined behavior.
This patch ensures the ordering of the two function calls by pulling
them both out of the goroutine and separating them with a WaitGroup.
Co-authored-by: James Shubin <james@shubin.ca>
Test to ensure that SocketSet is nonblocking and will close when
SocketSet.Shutdown() is called. Create a SocketSet that will never
receive any data and leave it running in a goroutine with a WaitGroup
for a second. If Shutdown is working correctly, the goroutine will be
terminated after the timer expires.
Adds a CPU count fact, that can be used to determine how many CPUs are
presently on the machine and ready for use (online). We get this by
reading from a netlink socket to the kernel, and the kernel sends us
uevents when CPUs are added, removed, and brought online or offline.
Whenever one of these events are received, we look in sysfs to update
the fact's Stream with the number of online CPUs.
Add the ReceiveBytes, ReceiveNetlinkMessage, and ReceiveUEvent methods.
This is because not everything passed through a netlink socket cannot
reliably be parsed using the ParseNetLinkMessage function.
With the ReceiveUEvent method, we add support for "uevent" kernel
events, which updates us about the state of devices currently on the
system. To make using this method easier, we add a UEvent struct, that
has the action (what event), Devpath (where the device lives in /proc or
/sysfs), and Subsystem (what subsystem this event belows to).
This continues the earlier patch that allowed resource names to be lists
of strings so that edges can now allow the same. This also includes a
new fancy test!
This greatly expands our test infra to allow us to drop in mcl tests and
look at their resource graph output. The only downside is that this only
runs the function engine once, so if the function graph would be
constantly changing over time, then this is not a good fit here.
I found a case where we had two missing unification rules. Now fixed in
the previous commits, and including this test to show I'm responsible.
I've added the same test in two locations for redundancy and as an
example.
I forgot to ensure that the type of the final expression matched the
type of each of the branches. It's rare, but possible for this to occur.
Luckily, this never would have caused a panic, because the func engine
would have caught the issue anyways, but it's still better we catch it
here first!
I forgot to include these two invariants which are occasionally
necessary, although in most cases they're necessary to prevent incorrect
code from getting past unification. In any case, they would have been
caught by the engine.
A clean re-write of this etcd code is needed, but until then, this
should hopefully workaround the occasional test failures. In practice I
don't think anyone has every hit this bug.
It's plausible that we send on a closed channel if we're running a back
poke and it tries to send a poke on something that has already closed.
If it detects this condition, it will exit.
Unfortunately, it's not clear if the wait group will protect this case,
but hopefully this will hold us until we can re-write the needed parts
of the engine.
Occasionally when a back poke happens downstream of an upstream vertex
which has already exited, it could get back poked, which would cause a
panic. This moves the deletion of the state struct until the entire
graph has completed so that it won't panic. It doesn't matter if a back
poke is lost, we're shutting down or pausing, and in this scenario it
can be lost.
Somewhere after the engine re-write we seem to have regressed and
converge early even if some resource is dirty. This adds an additional
timer so that we don't start the individual resource converged countdown
until our state is okay.
These weren't yet exposed in mcl. They're now available under the same
Meta namespace as the normal meta param structs. Even though they live
as a separate trait, they should be exposed together for a consistent
interface in mcl. If autoedge or autogroup ever grow additional params,
we can always add: `Meta:autoedge:something` to break it down further.
This adds a core looping construct by allowing a list of names to build
a resource. They'll all have the same parameters, but they'll
intelligently add the correct list of edges that they'd individually
create.
Constructs like these are one reason we do NOT have actual looping
functionality in the language, and it should stay that way.
Instead of adding complexity to the unification engine, we can add a
fake placeholder expression that is unreachable by the AST, but used for
unification so that we can ensure a "wrap" invariant has some contents.
Ideally we'd improve the unification engine, but we'll leave that for
the future, and it's easy to revert this one commit in the future.
Golang decided to change the GOCACHE behaviour in newer versions of `go
test`. This changes our tests to use the new approach.
For users using a local `.envrc`, you might want to add:
GOFLAGS="-count=1"
Which is supposed to fix this problem for local tests.
More information is available in: https://github.com/golang/go/issues/29378
This signals to an interested consumer that two or more compatible
resources can be merged safely. This is so that we can avoid the
"duplicate resource" design problem that puppet had.
To test this, you can run:
./mgmt run --tmp-prefix lang --lang 'pkg "cowsay" { state => "installed", } pkg "cowsay" { state => "newest", }'
which should work.
