How-To: Save and get state

Use key value pairs to persist a state

Introduction

State management is one of the most common needs of any application: new or legacy, monolith or microservice. Dealing with different databases libraries, testing them, handling retries and faults can be time consuming and hard.

Dapr provides state management capabilities that include consistency and concurrency options. In this guide we’ll start of with the basics: Using the key/value state API to allow an application to save, get and delete state.

Pre-requisites

Step 1: Setup a state store

A state store component represents a resource that Dapr uses to communicate with a database.

For the purpose of this guide we’ll use a Redis state store, but any state store from the supported list will work.

When using dapr init in Standalone mode, the Dapr CLI automatically provisions a state store (Redis) and creates the relevant YAML in a components directory, which for Linux/MacOS is $HOME/.dapr/components and for Windows is %USERPROFILE%\.dapr\components

To optionally change the state store being used, replace the YAML file statestore.yaml under /components with the file of your choice.

To deploy this into a Kubernetes cluster, fill in the metadata connection details of your desired statestore component in the yaml below, save as statestore.yaml, and run kubectl apply -f statestore.yaml.

  1. apiVersion: dapr.io/v1alpha1
  2. kind: Component
  3. metadata:
  4.   name: statestore
  5.   namespace: default
  6. spec:
  7.   type: state.redis
  8.   version: v1
  9.   metadata:
  10.   - name: redisHost
  11.     value: localhost:6379
  12.   - name: redisPassword
  13.     value: ""

See the instructions here on how to setup different state stores on Kubernetes.

Step 2: Save and retrieve a single state

The following example shows how to a single key/value pair using the Dapr state building block.

Note

It is important to set an app-id, as the state keys are prefixed with this value. If you don’t set it one is generated for you at runtime, and the next time you run the command a new one will be generated and you will no longer be able to access previously saved state.

Begin by launching a Dapr sidecar:

  1. dapr run --app-id myapp --dapr-http-port 3500

Then in a separate terminal save a key/value pair into your statestore:

  1. curl -X POST -H "Content-Type: application/json" -d '{ "key": "key1", "value": "value1"}' http://localhost:3500/v1.0/state/statestore

Now get the state you just saved:

  1. curl http://localhost:3500/v1.0/state/statestore/key1

You can also restart your sidecar and try retrieving state again to see that state persists separate from the app.

Begin by launching a Dapr sidecar:

  1. dapr --app-id myapp --port 3500 run

Then in a separate terminal save a key/value pair into your statestore:

  1. Invoke-RestMethod -Method Post -ContentType 'application/json' -Body '{"key": "key1", "value": "value1"}' -Uri 'http://localhost:3500/v1.0/state/statestore'

Now get the state you just saved:

  1. Invoke-RestMethod -Uri 'http://localhost:3500/v1.0/state/statestore/key1'

You can also restart your sidecar and try retrieving state again to see that state persists separate from the app.

Save the following to a file named pythonState.py:

  1. from dapr.clients import DaprClient
  3. with DaprClient() as d:
  4.     d.save_state(store_name="statestore", key="myFirstKey", value="myFirstValue" )
  5.     print("State has been stored")
  7.     data = d.get_state(store_name="statestore", key="myFirstKey").data
  8.     print(f"Got value: {data}")

Once saved run the following command to launch a Dapr sidecar and run the Python application:

  1. dapr --app-id myapp run python pythonState.py

You should get an output similar to the following, which will show both the Dapr and app logs:

  1. == DAPR == time="2021-01-06T21:34:33.7970377-08:00" level=info msg="starting Dapr Runtime -- version 0.11.3 -- commit a1a8e11" app_id=Braidbald-Boot scope=dapr.runtime type=log ver=0.11.3
  2. == DAPR == time="2021-01-06T21:34:33.8040378-08:00" level=info msg="standalone mode configured" app_id=Braidbald-Boot scope=dapr.runtime type=log ver=0.11.3
  3. == DAPR == time="2021-01-06T21:34:33.8040378-08:00" level=info msg="app id: Braidbald-Boot" app_id=Braidbald-Boot scope=dapr.runtime type=log ver=0.11.3
  4. == DAPR == time="2021-01-06T21:34:33.9750400-08:00" level=info msg="component loaded. name: statestore, type: state.redis" app_id=Braidbald-Boot scope=dapr.runtime type=log ver=0.11.3
  5. == DAPR == time="2021-01-06T21:34:33.9760387-08:00" level=info msg="API gRPC server is running on port 51656" app_id=Braidbald-Boot scope=dapr.runtime type=log ver=0.11.3
  6. == DAPR == time="2021-01-06T21:34:33.9770372-08:00" level=info msg="dapr initialized. Status: Running. Init Elapsed 172.9994ms" app_id=Braidbald-Boot scope=dapr.
  8. Checking if Dapr sidecar is listening on GRPC port 51656
  9. Dapr sidecar is up and running.
  10. Updating metadata for app command: python pythonState.py
  11. You are up and running! Both Dapr and your app logs will appear here.
  13. == APP == State has been stored
  14. == APP == Got value: b'myFirstValue'

Save the following in state-example.php:

  1. <?php
  2. require_once __DIR__.'/vendor/autoload.php';
  4. $app = \Dapr\App::create();
  5. $app->run(function(\Dapr\State\StateManager $stateManager, \Psr\Log\LoggerInterface $logger) {
  6.     $stateManager->save_state(store_name: 'statestore', item: new \Dapr\State\StateItem(
  7.         key: 'myFirstKey',
  8.         value: 'myFirstValue' 
  9.     ));
  10.     $logger->alert('State has been stored');
  12.     $data = $stateManager->load_state(store_name: 'statestore', key: 'myFirstKey')->value;
  13.     $logger->alert("Got value: {data}", ['data' => $data]);
  14. });

Once saved run the following command to launch a Dapr sidecar and run the PHP application:

  1. dapr --app-id myapp run -- php state-example.php

You should get an output similar to the following, which will show both the Dapr and app logs:

  1. ✅  You're up and running! Both Dapr and your app logs will appear here.
  3. == APP == [2021-02-12T16:30:11.078777+01:00] APP.ALERT: State has been stored [] []
  5. == APP == [2021-02-12T16:30:11.082620+01:00] APP.ALERT: Got value: myFirstValue {"data":"myFirstValue"} []

Step 3: Delete state

The following example shows how to delete an item by using a key with the state management API:

With the same dapr instance running from above run:

  1. curl -X DELETE 'http://localhost:3500/v1.0/state/statestore/key1'

Try getting state again and note that no value is returned.

With the same dapr instance running from above run:

  1. Invoke-RestMethod -Method Delete -Uri 'http://localhost:3500/v1.0/state/statestore/key1'

Try getting state again and note that no value is returned.

Update pythonState.py with:

  1. from dapr.clients import DaprClient
  3. with DaprClient() as d:
  4.     d.save_state(store_name="statestore", key="key1", value="value1" )
  5.     print("State has been stored")
  7.     data = d.get_state(store_name="statestore", key="key1").data
  8.     print(f"Got value: {data}")
  10.     d.delete_state(store_name="statestore", key="key1")
  12.     data = d.get_state(store_name="statestore", key="key1").data
  13.     print(f"Got value after delete: {data}")

Now run your program with:

  1. dapr --app-id myapp run python pythonState.py

You should see an output similar to the following:

  1. Starting Dapr with id Yakchocolate-Lord. HTTP Port: 59457. gRPC Port: 59458
  3. == DAPR == time="2021-01-06T22:55:36.5570696-08:00" level=info msg="starting Dapr Runtime -- version 0.11.3 -- commit a1a8e11" app_id=Yakchocolate-Lord scope=dapr.runtime type=log ver=0.11.3
  4. == DAPR == time="2021-01-06T22:55:36.5690367-08:00" level=info msg="standalone mode configured" app_id=Yakchocolate-Lord scope=dapr.runtime type=log ver=0.11.3
  5. == DAPR == time="2021-01-06T22:55:36.7220140-08:00" level=info msg="component loaded. name: statestore, type: state.redis" app_id=Yakchocolate-Lord scope=dapr.runtime type=log ver=0.11.3
  6. == DAPR == time="2021-01-06T22:55:36.7230148-08:00" level=info msg="API gRPC server is running on port 59458" app_id=Yakchocolate-Lord scope=dapr.runtime type=log ver=0.11.3
  7. == DAPR == time="2021-01-06T22:55:36.7240207-08:00" level=info msg="dapr initialized. Status: Running. Init Elapsed 154.984ms" app_id=Yakchocolate-Lord scope=dapr.runtime type=log ver=0.11.3
  9. Checking if Dapr sidecar is listening on GRPC port 59458
  10. Dapr sidecar is up and running.
  11. Updating metadata for app command: python pythonState.py
  12. You're up and running! Both Dapr and your app logs will appear here.
  14. == APP == State has been stored
  15. == APP == Got value: b'value1'
  16. == APP == Got value after delete: b''

Update state-example.php with the following contents:

  1. <?php
  2. require_once __DIR__.'/vendor/autoload.php';
  4. $app = \Dapr\App::create();
  5. $app->run(function(\Dapr\State\StateManager $stateManager, \Psr\Log\LoggerInterface $logger) {
  6.     $stateManager->save_state(store_name: 'statestore', item: new \Dapr\State\StateItem(
  7.         key: 'myFirstKey',
  8.         value: 'myFirstValue' 
  9.     ));
  10.     $logger->alert('State has been stored');
  12.     $data = $stateManager->load_state(store_name: 'statestore', key: 'myFirstKey')->value;
  13.     $logger->alert("Got value: {data}", ['data' => $data]);
  15.     $stateManager->delete_keys(store_name: 'statestore', keys: ['myFirstKey']);
  16.     $data = $stateManager->load_state(store_name: 'statestore', key: 'myFirstKey')->value;
  17.     $logger->alert("Got value after delete: {data}", ['data' => $data]);
  18. });

Now run it with:

  1. dapr --app-id myapp run -- php state-example.php

You should see something similar the following output:

  1. ✅  You're up and running! Both Dapr and your app logs will appear here.
  3. == APP == [2021-02-12T16:38:00.839201+01:00] APP.ALERT: State has been stored [] []
  5. == APP == [2021-02-12T16:38:00.841997+01:00] APP.ALERT: Got value: myFirstValue {"data":"myFirstValue"} []
  7. == APP == [2021-02-12T16:38:00.845721+01:00] APP.ALERT: Got value after delete:  {"data":null} []

Step 4: Save and retrieve multiple states

Dapr also allows you to save and retrieve multiple states in the same call.

With the same dapr instance running from above save two key/value pairs into your statestore:

  1. curl -X POST -H "Content-Type: application/json" -d '[{ "key": "key1", "value": "value1"}, { "key": "key2", "value": "value2"}]' http://localhost:3500/v1.0/state/statestore

Now get the states you just saved:

  1. curl -X POST -H "Content-Type: application/json" -d '{"keys":["key1", "key2"]}' http://localhost:3500/v1.0/state/statestore/bulk

With the same dapr instance running from above save two key/value pairs into your statestore:

  1. Invoke-RestMethod -Method Post -ContentType 'application/json' -Body '[{ "key": "key1", "value": "value1"}, { "key": "key2", "value": "value2"}]' -Uri 'http://localhost:3500/v1.0/state/statestore'

Now get the states you just saved:

  1. Invoke-RestMethod -Method Post -ContentType 'application/json' -Body '{"keys":["key1", "key2"]}' -Uri 'http://localhost:3500/v1.0/state/statestore/bulk'

The StateItem object can be used to store multiple Dapr states with the save_states and get_states methods.

Update your pythonState.py file with the following code:

  1. from dapr.clients import DaprClient
  2. from dapr.clients.grpc._state import StateItem
  4. with DaprClient() as d:
  5.     s1 = StateItem(key="key1", value="value1")
  6.     s2 = StateItem(key="key2", value="value2")
  8.     d.save_bulk_state(store_name="statestore", states=[s1,s2])
  9.     print("States have been stored")
  11.     items = d.get_bulk_state(store_name="statestore", keys=["key1", "key2"]).items
  12.     print(f"Got items: {[i.data for i in items]}")

Now run your program with:

  1. dapr --app-id myapp run python pythonState.py

You should see an output similar to the following:

  1. == DAPR == time="2021-01-06T21:54:56.7262358-08:00" level=info msg="starting Dapr Runtime -- version 0.11.3 -- commit a1a8e11" app_id=Musesequoia-Sprite scope=dapr.runtime type=log ver=0.11.3
  2. == DAPR == time="2021-01-06T21:54:56.7401933-08:00" level=info msg="standalone mode configured" app_id=Musesequoia-Sprite scope=dapr.runtime type=log ver=0.11.3
  3. == DAPR == time="2021-01-06T21:54:56.8754240-08:00" level=info msg="Initialized name resolution to standalone" app_id=Musesequoia-Sprite scope=dapr.runtime type=log ver=0.11.3
  4. == DAPR == time="2021-01-06T21:54:56.8844248-08:00" level=info msg="component loaded. name: statestore, type: state.redis" app_id=Musesequoia-Sprite scope=dapr.runtime type=log ver=0.11.3
  5. == DAPR == time="2021-01-06T21:54:56.8854273-08:00" level=info msg="API gRPC server is running on port 60614" app_id=Musesequoia-Sprite scope=dapr.runtime type=log ver=0.11.3
  6. == DAPR == time="2021-01-06T21:54:56.8854273-08:00" level=info msg="dapr initialized. Status: Running. Init Elapsed 145.234ms" app_id=Musesequoia-Sprite scope=dapr.runtime type=log ver=0.11.3
  8. Checking if Dapr sidecar is listening on GRPC port 60614
  9. Dapr sidecar is up and running.
  10. Updating metadata for app command: python pythonState.py
  11. You're up and running! Both Dapr and your app logs will appear here.
  13. == APP == States have been stored
  14. == APP == Got items: [b'value1', b'value2']

To batch load and save state with PHP, just create a “Plain Ole’ PHP Object” (POPO) and annotate it with the StateStore annotation.

Update the state-example.php file:

  1. <?php
  3. require_once __DIR__.'/vendor/autoload.php';
  5. #[\Dapr\State\Attributes\StateStore('statestore', \Dapr\consistency\EventualLastWrite::class)]
  6. class MyState {
  7.     public string $key1 = 'value1';
  8.     public string $key2 = 'value2';
  9. }
  11. $app = \Dapr\App::create();
  12. $app->run(function(\Dapr\State\StateManager $stateManager, \Psr\Log\LoggerInterface $logger) {
  13.     $obj = new MyState();
  14.     $stateManager->save_object(item: $obj);
  15.     $logger->alert('States have been stored');
  17.     $stateManager->load_object(into: $obj);
  18.     $logger->alert("Got value: {data}", ['data' => $obj]);
  19. });

Run the app:

  1. dapr --app-id myapp run -- php state-example.php

And see the following output:

  1. ✅  You're up and running! Both Dapr and your app logs will appear here.
  3. == APP == [2021-02-12T16:55:02.913801+01:00] APP.ALERT: States have been stored [] []
  5. == APP == [2021-02-12T16:55:02.917850+01:00] APP.ALERT: Got value: [object MyState] {"data":{"MyState":{"key1":"value1","key2":"value2"}}} []

Step 5: Perform state transactions

Note

State transactions require a state store that supports multi-item transactions. Visit the supported state stores page page for a full list. Note that the default Redis container created in a self-hosted environment supports them.

With the same dapr instance running from above perform two state transactions:

  1. curl -X POST -H "Content-Type: application/json" -d '{"operations": [{"operation":"upsert", "request": {"key": "key1", "value": "newValue1"}}, {"operation":"delete", "request": {"key": "key2"}}]}' http://localhost:3500/v1.0/state/statestore/transaction

Now see the results of your state transactions:

  1. curl -X POST -H "Content-Type: application/json" -d '{"keys":["key1", "key2"]}' http://localhost:3500/v1.0/state/statestore/bulk

With the same dapr instance running from above save two key/value pairs into your statestore:

  1. Invoke-RestMethod -Method Post -ContentType 'application/json' -Body '{"operations": [{"operation":"upsert", "request": {"key": "key1", "value": "newValue1"}}, {"operation":"delete", "request": {"key": "key2"}}]}' -Uri 'http://localhost:3500/v1.0/state/statestore'

Now see the results of your state transactions:

  1. Invoke-RestMethod -Method Post -ContentType 'application/json' -Body '{"keys":["key1", "key2"]}' -Uri 'http://localhost:3500/v1.0/state/statestore/bulk'

The TransactionalStateOperation can perform a state transaction if your state stores need to be transactional.

Update your pythonState.py file with the following code:

  1. from dapr.clients import DaprClient
  2. from dapr.clients.grpc._state import StateItem
  3. from dapr.clients.grpc._request import TransactionalStateOperation, TransactionOperationType
  5. with DaprClient() as d:
  6.     s1 = StateItem(key="key1", value="value1")
  7.     s2 = StateItem(key="key2", value="value2")
  9.     d.save_bulk_state(store_name="statestore", states=[s1,s2])
  10.     print("States have been stored")
  12.     d.execute_state_transaction(
  13.         store_name="statestore",
  14.         operations=[
  15.             TransactionalStateOperation(key="key1", data="newValue1", operation_type=TransactionOperationType.upsert),
  16.             TransactionalStateOperation(key="key2", data="value2", operation_type=TransactionOperationType.delete)
  17.         ]
  18.     )
  19.     print("State transactions have been completed")
  21.     items = d.get_bulk_state(store_name="statestore", keys=["key1", "key2"]).items
  22.     print(f"Got items: {[i.data for i in items]}")

Now run your program with:

  1. dapr run python pythonState.py

You should see an output similar to the following:

  1. Starting Dapr with id Singerchecker-Player. HTTP Port: 59533. gRPC Port: 59534
  2. == DAPR == time="2021-01-06T22:18:14.1246721-08:00" level=info msg="starting Dapr Runtime -- version 0.11.3 -- commit a1a8e11" app_id=Singerchecker-Player scope=dapr.runtime type=log ver=0.11.3
  3. == DAPR == time="2021-01-06T22:18:14.1346254-08:00" level=info msg="standalone mode configured" app_id=Singerchecker-Player scope=dapr.runtime type=log ver=0.11.3
  4. == DAPR == time="2021-01-06T22:18:14.2747063-08:00" level=info msg="component loaded. name: statestore, type: state.redis" app_id=Singerchecker-Player scope=dapr.runtime type=log ver=0.11.3
  5. == DAPR == time="2021-01-06T22:18:14.2757062-08:00" level=info msg="API gRPC server is running on port 59534" app_id=Singerchecker-Player scope=dapr.runtime type=log ver=0.11.3
  6. == DAPR == time="2021-01-06T22:18:14.2767059-08:00" level=info msg="dapr initialized. Status: Running. Init Elapsed 142.0805ms" app_id=Singerchecker-Player scope=dapr.runtime type=log ver=0.11.3
  8. Checking if Dapr sidecar is listening on GRPC port 59534
  9. Dapr sidecar is up and running.
  10. Updating metadata for app command: python pythonState.py
  11. You're up and running! Both Dapr and your app logs will appear here.
  13. == APP == State transactions have been completed
  14. == APP == Got items: [b'value1', b'']

Transactional state is supported by extending TransactionalState base object which hooks into your object via setters and getters to provide a transaction. Before you created your own transactional object, but now you’ll ask the Dependency Injection framework to build one for you.

Modify the state-example.php file again:

  1. <?php
  3. require_once __DIR__.'/vendor/autoload.php';
  5. #[\Dapr\State\Attributes\StateStore('statestore', \Dapr\consistency\EventualLastWrite::class)]
  6. class MyState extends \Dapr\State\TransactionalState {
  7.     public string $key1 = 'value1';
  8.     public string $key2 = 'value2';
  9. }
  11. $app = \Dapr\App::create();
  12. $app->run(function(MyState $obj, \Psr\Log\LoggerInterface $logger, \Dapr\State\StateManager $stateManager) {
  13.     $obj->begin();
  14.     $obj->key1 = 'hello world';
  15.     $obj->key2 = 'value3';
  16.     $obj->commit();
  17.     $logger->alert('Transaction committed!');
  19.     // begin a new transaction which reloads from the store
  20.     $obj->begin();
  21.     $logger->alert("Got value: {key1}, {key2}", ['key1' => $obj->key1, 'key2' => $obj->key2]);
  22. });

Run the application:

  1. dapr --app-id myapp run -- php state-example.php

Observe the following output:

  1. ✅  You're up and running! Both Dapr and your app logs will appear here.
  3. == APP == [2021-02-12T17:10:06.837110+01:00] APP.ALERT: Transaction committed! [] []
  5. == APP == [2021-02-12T17:10:06.840857+01:00] APP.ALERT: Got value: hello world, value3 {"key1":"hello world","key2":"value3"} []

Next steps

Last modified January 1, 0001