azure_functions

Lithops runtime for Azure Functions

The runtime is the place where your functions are executed.

The default runtime is created the first time you execute a function. Lithops automatically detects the Python version of your environment and deploys the default runtime based on it.

Currently, Azure Functions supports Python 3.6, 3.7, 3.8 and 3.9, and it provides the following default runtimes with some packages already preinstalled:

Runtime name	Python version	Packages included
lithops-runtime-v36	3.6	list of packages
lithops-runtime-v38	3.7	list of packages
lithops-runtime-v38	3.8	list of packages
lithops-runtime-v38	3.9	list of packages

Lithops default runtimes are also ship with the following packages:

azure-functions
azure-storage-blob
azure-storage-queue
pika
flask
gevent
redis
requests
PyYAML
kubernetes
numpy
cloudpickle
ps-mem
tblib

To run a function with the default runtime you don't need to specify anything in the code, since everything is managed internally by Lithops:

import lithops

def my_function(x):
    return x + 7

fexec = lithops.FunctionExecutor()
fexec.call_async(my_function, 3)
result = lithops.get_result()

• Note that Azure Functions does not allow to set a specific memory size for the runtimes, so the parameter runtime_memory won't take effect.

Custom runtime

Build your own Lithops runtime for Azure Functions

If you need some Python modules which are not included in the default runtime, it is possible to build your own Lithops runtime with all of them.

To build your own runtime, you have to collect all necessary modules in a requirements.txt file. For example, if you want to add the modules numpy and matplotlib to our runtime, since they are not provided in the default runtime.

First, we need to extend the requirements.txt file provided with Lithops with all the modules we need. For our example, the requirements.txt should contain the following modules (note that we added numpy and matplotlib at the end):

azure-functions
azure-storage-blob
azure-storage-queue
pika
flask
gevent
redis
requests
PyYAML
kubernetes
cloudpickle
ps-mem
tblib
numpy
matplotlib

IMPORTANT: Note that the runtime is built using your local machine, and some libraries, like Numpy, compile some C code based on the Operating System you are using. Azure functions run on a Linux machine, this mean that if you use MacOS or Windows for building the runtime, those libraries that compiled C code cannot be executed from within the function. In this case, you must use a Linux machine for building the runtime.

Then, we will build the runtime, specifying the modified requirements.txt file and a runtime name:

$ lithops runtime build -b azure_functions -f requirements.txt matplotlib-runtime 

This command will built a runtime called matplotlib-runtime in your local machine. Then, to deploy the runtime to the available Azure Functions runtimes, execute:

$ lithops runtime deploy -b azure_functions -s azure_storage matplotlib-runtime 

Finally, we can specify this new runtime when creating a Lithops Function Executor:

import lithops

def test():
    import matplotlib
    return repr(matplotlib)

lith = lithops.FunctionExecutor(runtime='matplotlib-runtime')
lith.call_async(test, data=())
res = lith.get_result()
print(res)

Note that both the client and the runtime must have the same python version. If you are running Lithops, for example, with Python 3.8, the matplotlib-runtime will be a Python 3.8 runtime with the extra modules specified installed.