Plugin Reference
The cljs-lambda plugin enables the declaration of Lambda-deployable functions
from Leiningen project files.
In decreasing order of precedence, configuration parameters will be looked for in the following places:
- Command line
- The active Lambda function's map in your project file (for function-specific commands)
- Project file's top-level
cljs-lambdaconfig map
Command line options are specified as keyword literals, e.g.
$ lein cljs-lambda deploy :region us-west-2There's a distinction between commands which operate on functions defined in
your project (e.g. deploy, update-configs) and general purpose operations
which can also be used to deal with foreign Lambda functions (e.g. invoke,
alias). The latter do not read values from your function's configuration.
There's a few options which apply to all, or most commands:
Names the AWS CLI profile from which credentials'll be retrieve.
Defaults to the default region of the active AWS profile
Silences debug console messages, but preserves the eventual command output, if any.
Controls the level of parallelism for deploy and update-config. Network is
likely to be the limiting factor for deploy. Defaults to 5.
{:cljs-build-id "cljs-lambda-example"
:compiler {:inputs ["src"]
:options {:output-to ... :optimizations ... ...}}
:aws-profile "cljs-lambda-profile"
:defaults {:role "arn:aws:iam::151963828411:role/..."
:memory-size 512
:runtime :nodejs}
:env {:capture #{"EXACT_MATCH" #"PARTIAL_"}
:set {"CUSTOM_VAR" "value"}}
:resource-dirs ["config"]
:region us-west-2
:functions
[{:name "dog-bark"
:invoke cljs-lambda-example.dog/bark
:region us-east-1
:timeout 30}
{:name "cat-meow"
:invoke cljs-lambda-example.cat/meow
:create false
:role "arn:aws:iam::151963828411:role/..."
:description "Semi-pro"
:vpc {:subnets ["subnet-xyz"]
:security-groups ["sg-abc" "sg-def"]}}]}(excludes :compiler)
Names the cljsbuild entry which describes how your project ought to be built
for the purposes of Lambda deployment. Defaults to the first build, in the absence of :compile.
(excludes :cljs-build-id)
:inputs & :options to pass to the Clojurescript compiler.
A sequence of directories, each appearing at the top-level in the resulting zip file.
:capture can be used to make available at runtime (on Lambda) portions of the
compile/deploy-time environment. The values of matching variables will be set
in process.env before your project is loaded, from where they may be retrieved
via cljs-lambda.context/env (or via Node's API, though env'll help out w/
testing)
:set incorporates into the runtime environment arbitrary env values,
overriding any conflicting names from :capture.
IAM execution role identifier. A suitable placeholder can be created with lein cljs-lambda default-iam-role. While required, it (and any of the other keys) can be supplied as defaults via e.g. :defaults {:role ...}` in the topmost map,
as above.
Virtual Private Cloud configuration. You have to provide at least one subnet and at least one security group. Subnets and security groups must be from the same VPC.
VPC is supported since 0.6.3-SNAPSHOT.
Determines whether a create-function Lambda command will be issued if an
attempt is made to deploy a not-yet-existing Lambda function. Defaults to
true.
$ lein cljs-lambda build
Used by invoke, but may be used directly to produce zip files
for manual uploading (e.g. via S3, if the file is over 50MB in size, per Lambda
restrictions). Writes a file in the
$ lein cljs-lambda deploy [function-names] [:publish :alias ...]
build, update-config (if changed), then deploy either all functions in the
project file, or the named subset. publish (reify deployment) and alias
(tag version) are only applicable for single-function deploys, and will
atomically publish and (optionally) name the version of the function being
deployed. Any config parameters allowed in a function's config map are accepted
from the command line.
If :publish is specified and :alias is not, outputs the published version
number.
$ lein cljs-lambda deploy dog-bark :publish :alias cujo :memory-size 512
$ lein cljs-lambda deploy :parallel 1 :region us-west-2 :quiet
$ lein cljs-lambda update-config [function-names] [...]
Sync function configuration (handler, role, description, timeout) for
either all functions in the project file, or the named subset. Any config
parameters allowed in a function's config map are accepted from the command
line, and will apply to all functions being deployed.
Avoids writing if the local and remote configurations are identical.
$ lein cljs-lambda update-config cat-meow dog-bark :timeout 16 :parallel
$ lein cljs-lambda invoke function-name [payload] [:region :aws-profile :version]
Remotely invokes the named function (either function name or full/partial ARN),
which needn't exist in the effective project file. :version (or the
alternatively named :qualifier) optionally identify either a version number or
alias. payload is a literal JSON string.
Outputs the function's result to stdout, with remote log messages printed to stderr.
$ lein cljs-lambda invoke echo '{"hello" [1,2,3]}' :version production
$ lein cljs-lambda invoke echo:production :region us-west-2
$ lein cljs-lambda default-iam-role [:aws-profile]
Creates a minimally permissive Lambda execution role in the account
corresponding to the default or named AWS CLI profile, and modifies
project.clj to specify it as the function default.
$ lein cljs-lambda alias alias-name fn-name at-version [:region aws:profile]
Creates a tagged version for an already published Lambda function. The atomic
behaviour of invoke & :alias/:publish is probably a better idea.
$ lein cljs-lambda alias cujo dog-bark \$LATEST
$ lein cljs-lambda deploy dog-bark :publish
1
$ lein cljs-lambda alias cujo dog-bark 1