I tried to re-generate the Protobuf/gRPC files, but generate.sh seems to
have assumed I manually built the Dockerfile ahead-of-time (unless I'm
missing something -- very possible). Unfortunately, when I went to do
that, I ended up with a handful of errors, most of them due to
differences in versions. (We probably want to think about pinning them.)
To remedy both issues, I've fixed up the Dockerfile so that it works for
me at least, and added a build step to the front of generate.sh.
A critical part of the partial update protocol is to return a structured
error when a resource is successfully created, but fails to initialize.
This structured error contains the properties of the
partially-initialized resource, and instructs the engine to halt.
Most languages implement this by attaching "details" to the error, i.e.,
an arbitrary proto message attached to the error. The JavaScript
implementation is not mature enough to include all the facilities
required to use this, so here we must add a `Status` message, which
protobuf requires as part of its structure for returning details.
This change includes a bunch of refactorings I made in prep for
doing refresh (first, the command, see pulumi/pulumi#1081):
* The primary change is to change the way the engine's core update
functionality works with respect to deploy.Source. This is the
way we can plug in new sources of resource information during
planning (and, soon, diffing). The way I intend to model refresh
is by having a new kind of source, deploy.RefreshSource, which
will let us do virtually everything about an update/diff the same
way with refreshes, which avoid otherwise duplicative effort.
This includes changing the planOptions (nee deployOptions) to
take a new SourceFunc callback, which is responsible for creating
a source specific to the kind of plan being requested.
Preview, Update, and Destroy now are primarily differentiated by
the kind of deploy.Source that they return, rather than sprinkling
things like `if Destroying` throughout. This tidies up some logic
and, more importantly, gives us precisely the refresh hook we need.
* Originally, we used the deploy.NullSource for Destroy operations.
This simply returns nothing, which is how Destroy works. For some
reason, we were no longer doing this, and instead had some
`if Destroying` cases sprinkled throughout the deploy.EvalSource.
I think this is a vestige of some old way we did configuration, at
least judging by a comment, which is apparently no longer relevant.
* Move diff and diff-printing logic within the engine into its own
pkg/engine/diff.go file, to prepare for upcoming work.
* I keep noticing benign diffs anytime I regenerate protobufs. I
suspect this is because we're also on different versions. I changed
generate.sh to also dump the version into grpc_version.txt. At
least we can understand where the diffs are coming from, decide
whether to take them (i.e., a newer version), and ensure that as
a team we are monotonically increasing, and not going backwards.
* I also tidied up some tiny things I noticed while in there, like
comments, incorrect types, lint suppressions, and so on.
This change includes a lot more functionality. Enough to actually
run the webserver-py example through previews, updates, and destroys!
* Actually wire up the gRPC connections to the engine/monitor.
* Move the Node.js and Python generated Protobuf/gRPC files underneath
the actual SDK directories to simplify this generally. No more
copying during `make` and, in fact, this was required to give a smoother
experience with good packages/modules for the Python's SDK development.
* Build the Python egg during `make build`.
* Add support for program stacks. Just like with the Node.js runtime,
we will auto-parent any resources without explicit parents to a single
top-level resource component.
* Add support for component resource output properties.
* Add get_project() and get_stack() functions for retrieving the current
project and stack names.
* Properly use UNKNOWN sentinels.
* Add a set_outputs() function on Resource. This is defined by the
code-generator and allows custom logic for output property setting.
This is cleaner than the way we do this in Node.js, and gives us a
way to ensure that output properties are "real" properties, complete
with member documentation. This also gives us a hook to perform
name demangling, which the code-generator typically controls anyway.
* Add package dependencies to setuptools.py and requirements.txt.
This is the initial step towards redefining Lumi as a library that runs
atop vanilla Node.js/V8, rather than as its own runtime.
This change is woefully incomplete but this includes some of the more
stable pieces of my current work-in-progress.
The new structure is that within the sdk/ directory we will have a client
library per language. This client library contains the object model for
Lumi (resources, properties, assets, config, etc), in addition to the
"language runtime host" components required to interoperate with the
Lumi resource monitor. This resource monitor is effectively what we call
"Lumi" today, in that it's the thing orchestrating plans and deployments.
Inside the sdk/ directory, you will find nodejs/, the Node.js client
library, alongside proto/, the definitions for RPC interop between the
different pieces of the system. This includes existing RPC definitions
for resource providers, etc., in addition to the new ones for hosting
different language runtimes from within Lumi.
These new interfaces are surprisingly simple. There is effectively a
bidirectional RPC channel between the Lumi resource monitor, represented
by the lumirpc.ResourceMonitor interface, and each language runtime,
represented by the lumirpc.LanguageRuntime interface.
The overall orchestration goes as follows:
1) Lumi decides it needs to run a program written in language X, so
it dynamically loads the language runtime plugin for language X.
2) Lumi passes that runtime a loopback address to its ResourceMonitor
service, while language X will publish a connection back to its
LanguageRuntime service, which Lumi will talk to.
3) Lumi then invokes LanguageRuntime.Run, passing information like
the desired working directory, program name, arguments, and optional
configuration variables to make available to the program.
4) The language X runtime receives this, unpacks it and sets up the
necessary context, and then invokes the program. The program then
calls into Lumi object model abstractions that internally communicate
back to Lumi using the ResourceMonitor interface.
5) The key here is ResourceMonitor.NewResource, which Lumi uses to
serialize state about newly allocated resources. Lumi receives these
and registers them as part of the plan, doing the usual diffing, etc.,
to decide how to proceed. This interface is perhaps one of the
most subtle parts of the new design, as it necessitates the use of
promises internally to allow parallel evaluation of the resource plan,
letting dataflow determine the available concurrency.
6) The program exits, and Lumi continues on its merry way. If the program
fails, the RunResponse will include information about the failure.
Due to (5), all properties on resources are now instances of a new
Property<T> type. A Property<T> is just a thin wrapper over a T, but it
encodes the special properties of Lumi resource properties. Namely, it
is possible to create one out of a T, other Property<T>, Promise<T>, or
to freshly allocate one. In all cases, the Property<T> does not "settle"
until its final state is known. This cannot occur before the deployment
actually completes, and so in general it's not safe to depend on concrete
resolutions of values (unlike ordinary Promise<T>s which are usually
expected to resolve). As a result, all derived computations are meant to
use the `then` function (as in `someValue.then(v => v+x)`).
Although this change includes tests that may be run in isolation to test
the various RPC interactions, we are nowhere near finished. The remaining
work primarily boils down to three things:
1) Wiring all of this up to the Lumi code.
2) Fixing the handful of known loose ends required to make this work,
primarily around the serialization of properties (waiting on
unresolved ones, serializing assets properly, etc).
3) Implementing lambda closure serialization as a native extension.
This ongoing work is part of pulumi/pulumi-fabric#311.
This change adds basic support for discovering, loading, binding to,
and invoking RPC methods on, resource provider plugins.
In a nutshell, we add a new context object that will share cached
state such as loaded plugins and connections to them. It will be
a policy decision in server scenarios how much state to share and
between whom. This context also controls per-resource context
allocation, which in the future will allow us to perform structured
cancellation and teardown amongst entire groups of requests.
Plugins are loaded based on their name, and can be found in one of
two ways: either simply by having them on your path (with a name of
"mu-ressrv-<pkg>", where "<pkg>" is the resource package name with
any "/"s replaced with "_"s); or by placing them in the standard
library installation location, which need not be on the path for this
to work (since we know precisely where to look).
If we find a protocol, we will load it as a child process.
The protocol for plugins is that they will choose a port on their
own -- to eliminate races that'd be involved should Mu attempt to
pre-pick one for them -- and then write that out as the first line
to STDOUT (terminated by a "\n"). This is the only STDERR/STDOUT
that Mu cares about; from there, the plugin is free to write all it
pleases (e.g., for logging, debugging purposes, etc).
Afterwards, we then bind our gRPC connection to that port, and create
a typed resource provider client. The CRUD operations that get driven
by plan application are then simple wrappers atop the underlying gRPC
calls. For now, we interpret all errors as catastrophic; in the near
future, we will probably want to introduce a "structured error"
mechanism in the gRPC interface for "transactional errors"; that is,
errors for which the server was able to recover to a safe checkpoint,
which can be interpreted as ResourceOK rather than ResourceUnknown.
After a bit more thinking, we will create new SDK packages for each
of the languages we wish to support writing resource providers in.
This is where the RPC goo will live, so I have created a new sdk/
directory, moved the Protobuf/gRPC definitions underneath sdk/proto/,
and put the generated code into sdk/go/ and sdk/js/.