Skip to main content
Version: 10.x (Current)

Real-Time Updater Common Configuration

Real-Time Updater is the service in charge of keeping up to date the projections with the data sent by the connected system.
Optionally, the service can generate several events so that your services can consume these events and react when projections are updated.

For having an overview of the features of the Real-Time Updater, you can go here.
Here below, instead, all the configurations the service accepts are explained.

Environment variables

NameRequiredDescriptionDefault value
LOG_LEVELtruedefines the level of the logger -
MONGODB_URLtruedefines the mongodb URL to contact -
PROJECTIONS_DATABASE_NAMEtruedefines the name of the projections' database -
PROJECTIONS_CHANGES_COLLECTION_NAMEtruedefines the name of the projections changes collection -
PROJECTIONS_CHANGES_ENABLEDfalsedefines whether you want to generate projections changestrue
LCtruedefines the lc39 HTTP port-
STRATEGIES_MAX_EXEC_TIME_MStruedefines the maximum time for which a strategy is executed-
KAFKA_BROKERStruedefines the Kafka brokers-
KAFKA_GROUP_IDtruedefines the Kafka group id (it is suggested to use a syntax like {tenant}.{environment}.{projectName}.{system}.real-time-updater)-
KAFKA_SASL_USERNAMEtruedefines the Kafka SASL username-
KAFKA_SASL_PASSWORDtruedefines the Kafka SASL password-
LIVENESS_INTERVAL_MStruedefines the liveness interval in milliseconds-
INVARIANT_TOPIC_MAPtruedefines an object that maps the topic to the projection-
KAFKA_USE_LATEST_DEQUEUE_STRATEGYfalsedefines latest dequeue strategy or not-
KAFKA_ADAPTER_FOLDERfalsedefines the path to the Kafka adapter folder-
CAST_FUNCTIONS_FOLDERfalsedefines the path to the cast-functions folder-
MAP_TABLE_FOLDERfalsedefines the path to the map table folder-
STRATEGIES_FOLDERfalsedefines the path to the strategies' folder-
KAFKA_SASL_MECHANISMfalsedefines the authentication mechanism. It can be one of: `plain`, `scram-sha-256` or `scram-sha-512`.plain
USE_UPSERTfalsedefines whether to use upsert or not when performing insert and update operations.true
KAFKA_MESSAGE_ADAPTERfalsedefines which Kafka message adapter to use. Its value can be either `basic`, `golden-gate` or `custom`. This value can be changed only in the related System of Records, on the _Projections_ page. Any manual update from the Environment Variables table will be loss when saving. Further details on the Kafka Adapters: Kafka messages format paragraph.basic
KAFKA_PROJECTION_CHANGES_FOLDERfalsepath where has been mounted the `kafkaProjectionChanges.json` configuration (v3.4.0 or above).-
GENERATE_KAFKA_PROJECTION_CHANGESfalsedefines whether the projection changes have to be sent to Kafka too or not. (v3.4.0 or above).false
KAFKA_CONSUMER_MAX_WAIT_TIMEfalsedefines the maximum waiting time of Kafka Consumer for new data in batch.500
COMMIT_MESSAGE_LOGGING_INTERVALfalsespecify the interval in ms of logging the info that messages have been committed.3000
KAFKA_CONNECTION_TIMEOUT_MSfalseTime in milliseconds to wait for a successful connection.10000
KAFKA_SESSION_TIMEOUT_MSfalseTimeout in milliseconds used to detect failures.30000
KAFKA_HEARTBEAT_INTERVAL_MSfalseThe expected time in milliseconds between heartbeats to the consumer coordinator.3000
FORCE_CHECK_ON_OFFSETfalseForce check that incoming message has offset greater or equal than the one of the projection to update.true
KAFKA_PROJECTION_UPDATES_FOLDERfalsepath to the folder that contains the file `kafkaProjectionUpdates.json`, containing configurations of the topic where to send the updates to, mapped to each projection. (v5.3.0 or above).-
GENERATE_KAFKA_PROJECTION_UPDATESfalsedefines whether the realtime updater should send a message of update every time it writes the projection to Mongo.false
CA_CERT_PATHfalsethe path to the CA certificate, which should include the file name as well, e.g. `/home/my-ca.pem`-
SYSTEM_OF_RECORDStruethe name of the system of records associated to the Real Time Updater-
PAUSE_TOPIC_CONSUMPTION_ON_ERRORfalseIf set to true, in case of an error while consuming an ingestion message, the service will pause the topic's consumption while keep consuming the other ones. More info on the feature herefalse

How data is managed on MongoDB

Projection Deletion

When the Real-Time Updater deletes a document, it actually makes a virtual delete instead of a real document deletion. (the __STATE__ field is set to DELETED)

Upsert

When performing Insert and Update operations, Real-Time Updater will perform an upsert operation as default. This means that if the document matching the key of the message does not exist, it will be created; otherwise, if it already exists, it will just be updated.

If you want to change this behavior, you can set the environment variable USE_UPSERT of the Real-Time Updater to false. Doing so, for the Insert operation the service will fail to insert documents that already exist, causing the service to stop. For the Update operation it will fail when trying to update non-existing documents, causing the service to stop.

Configuration files

The Real-Time Updater accepts the following configurations:

Kafka Adapters: Kafka messages format

In the Fast Data architecture CDC, iPaaS, APIs and sFTP publish messages on Kafka topic to capture change events. However, these messages could be written in different formats. The purpose of the Kafka adapter is allowing the correct reading of these messages in order to be properly consumed by the Real-Time Updater.

Since a Real Time Updater is connected to one System of Records, when creating a new System you will be asked to select the type of message adapter you want to use, between one of the following three options (further details in the paragraphs below):

When the System of Records has been created, saving the configuration will automatically create a new Real Time Updater service, which can be viewed in the Microservices section of the console.

In case you need to change the type of adapter to use in the Real Time Updater, you can easily do that from the System of Records page in the Projection section of the console: after selecting your System (the same one associated to the Real Time Updater), click on the Real Time Updater tab, then select the adapter type you want to use.

Another option that you should be aware of when thinking about the format of your Kafka messages is the "upsert" or "insert". By default, the real-time-updater will perform upsert operations, but you can optionally decide to perform inserts that will fail if the document already exists, instead of updating it.

IBM InfoSphere Data Replication for DB2

It's the default one, it is based on the IBM InfoSphere Data Replication engine.

The message should include the following properties:

  • timestamp: a stringified integer greater than zero. This integer has to be a valid timestamp.
  • key: it has to be a stringified object containing the primary key of the projection, if value also contains the primary key of the projection this field can be an empty string.
  • value: it is null if it's a delete operation, otherwise it contains the data of the projection.
  • offset: it is the offset of the Kafka message.

Example of a delete operation

key: `{"USER_ID": 123, "FISCAL_CODE": "ABCDEF12B02M100O"}`
value: null
timestamp: '1234556789'
offset: '100'

Example of an upsert:

key: `{"USER_ID": 123, "FISCAL_CODE": "ABCDEF12B02M100O"}`
value: `{"NAME": 456}`
timestamp: '1234556789'
offset: '100'

These are the only fields needed to configure correctly the message adapter. For more details and further explanations, you can read the documentation page about the supported JSON format.

danger

This Kafka message format does not support a Primary Key update. For additional information, please check the Primary Key update section.

Oracle GoldenGate

This Kafka Message Adapter has been created to have a format supported by Oracle Golden Gate.

In this Golden Gate adapter, we expect that the message includes data as explained in the JSON Formatter page of the official documentation.

The following is an example of value for an insert operation:

{
'table': 'MY_TABLE',
'op_type': 'I',
'op_ts': '2021-02-19 16:03:27.000000',
'current_ts': '2021-02-19T17:03:32.818003',
'pos': '00000000650028162190',
'after': {
'USER_ID': 123,
'FISCAL_CODE': 'the-fiscal-code-123',
'COINS': 300000000,
},
}

Custom

If you have Kafka Messages that do not match one of the formats above, you can create your own custom adapter for the messages.

To make this work, you need to create a Custom Kafka Message Adapter inside Real Time Updater section of the related System of Records. The adapter must be a javascript function that converts Kafka messages as received from the real-time updater to an object with a specific structure. This function must receives as arguments the Kafka message and the list of primary keys of the projection, and must return an object with the following properties:

  • offset: the offset of the Kafka message
  • timestampDate: an instance of Date of the timestamp of the Kafka message.
  • keyObject: an object containing the primary keys of the projection. It is used to know which projection document needs to be updated with the changes set in the value.
  • value: the data values of the projection, or null
  • operation: optional value that indicates the type of operation (either I for insert, U for update, or D for delete). It is not needed if you are using an upsert on insert logic (the default one), while it is required if you want to differentiate between insert and update messages.
  • before: optional value that indicates the data values before the operation execution
  • after: optional value that indicates the data values after the operation execution
  • operationPosition: optional value that indicates a positive integer, usually a timestamp, which ensures messages are processed in the correct order

If the value is null, it is a delete operation. The keyObject cannot be null.

note

To support a Primary Key update, the before, after and operationPosition fields should be included in the adapter. (Hint: if not present, a simple operationPosition value might be the Kafka message timestamp).

Inside configmap folder create your javascript file named kafkaMessageAdapter.js.

The file should export a simple function with the following signature:

module.exports = function kafkaMessageAdapter(kafkaMessage, primaryKeys, logger) {
const {
value: valueAsBuffer, // type Buffer
key: keyAsBuffer, // type Buffer
timestamp: timestampAsString, // type string
offset: offsetAsString, // type string
} = kafkaMessage

// your adapting logic

return {
keyObject: keyToReturn, // type object (NOT nullable)
value: valueToReturn, // type object (null or object)
timestampDate: new Date(parseInt(timestampAsString)), // type Date
offset: parseInt(offsetAsString), // type number
operation: operationToReturn, // type string (either I, U, or D)
}
}

The kafkaMessage argument is the Kafka message as received from the real-time-updater.
The fields value and key are of type Buffer, offset and timestamp are of type string.

The primaryKeys is an array of strings which are the primary keys of the projection whose topic is linked.

CAST_FUNCTION configurations

The mount path used for these configurations is: /home/node/app/configurations/castFunctionsFolder.
In this folder you have all the generated Cast Functions definitions. This configuration is read-only since you can configure it from its dedicated section of the Console.

MAP_TABLE configurations

The mount path used for these configurations is: /home/node/app/configurations/mapTableFolder.
Two mappings will be placed in this folder: one between cast functions and fields and another one between strategies and projections. This configuration is read-only since it's configured automatically based on the projections and strategies you configure from the Fast Data section of the Console.

Kafka Projection Changes configuration

Projection changes are saved on Mongo, but from version v3.4.0 and above, you can send them to Kafka as well.

This feature enables you to send the projection changes to a topic Kafka you want to. This is useful if you want to have a history of the projection changes thanks to the Kafka retention of messages. You can also make your own custom logic when a projection change occurs by setting a Kafka consumer attached to the topic Kafka you set.

info

This feature is available from the version v3.4.0 or above of the service

To do that, you need to set two environment variables:

  • KAFKA_PROJECTION_CHANGES_FOLDER: path where has been mounted the kafkaProjectionChanges.json configuration (v3.4.0 or above).
  • GENERATE_KAFKA_PROJECTION_CHANGES: defines whether the projection changes have to be sent to Kafka too or not. Default is false(v3.4.0 or above).

You have to create a configuration with the same path as the one defined by the environment variable KAFKA_PROJECTION_CHANGES_FOLDER. Then, you have to create a configuration file kafkaProjectionChanges.json inside that configuration. The configuration is a json file like the following one:

{
"MY_PROJECTION": {
"projectionChanges": {
"MY_SINGLE_VIEW": {
"strategy": "MY_STRATEGY",
"topic": "MY_TOPIC",
}
}
}
}

where:

  • MY_PROJECTION is the name of the collection whose topic has received the message from the CDC.
  • MY_SINGLE_VIEW is the single view that have to be updated
  • MY_STRATEGY is the strategy to be used to get the identifier of the Single View to update. It could be the name of a file or, in case an automatic strategy, the string __automatic__.
  • MY_TOPIC is the topic where the projection change need to be sent (for further information about the naming convention adopted for this topic, click here)

Example:

{
"registry-json": {
"projectionChanges": {
"sv_pointofsale": {
"strategy": "__automatic__",
"topic": "my-tenant.development.my-database.sv-pointofsale.sv-trigger",
}
}
},
"another-projection": {
"projectionChanges": {
"sv_customer": {
"strategy": "__fromFile__[myStrategy]",
"topic": "my-tenant.development.my-database.sv-customer.sv-trigger"
}
}
}
}

When a message about registry-json happens, the projection changes will be saved on MongoDB, and it will be sent to the Kafka topic my-tenant.development.my-database.sv-pointofsale.sv-trigger as well.

Custom Projection Changes Collection

You can choose to use a collection you have already created in the CRUD section.

In order to do that, your collection is supposed to have the following fields (apart from the default ones):

[
{"name":"type","type":"string","required":false,"nullable":false},
{"name":"changes","type":"Array_RawObject","required":false,"nullable":false},
{"name":"identifier","type":"RawObject","required":true,"nullable":false},
{"name":"doneAt","type":"Date","required":false,"nullable":false}
]

You also need to have the following additional indexes:

Add an index with name type_change_state, type normal, unique false.
You need to add the following index fields:

  • name changes.state, order ASCENDENT
  • name type, order ASCENDENT

Add another index with name type_identifier, type normal, unique true.
You need to add the following index fields:

  • name identifier, order ASCENDENT
  • name type, order ASCENDENT

After that, you need to set your collection as the one to be used by the Real-Time Updater. To do so, set the name of the collection you want to use as value of the PROJECTIONS_CHANGES_COLLECTION_NAME environment variable of the service.

note

To allow the Single View Creator to read from the Projection Changes, the collection name should also be set in the PROJECTIONS_CHANGES_COLLECTION environment variable of your Single View Creator service.

Kafka Projection Updates configuration

Whenever the real-time updater performs a change on Mongo on a projection, you can choose to send a message to a Kafka topic as well, containing information about the performed change and, if possible, the state of the projection before and after the change and the document ID of the document involved in the change.

info

This feature has been introduced since version v3.5.0 of the real time updater

To activate this feature you need to set the following environment variables:

  • KAFKA_PROJECTION_UPDATES_FOLDER: path to the folder that contains the file kafkaProjectionUpdates.json, containing configurations of the topic where to send the updates to, mapped to each projection.
  • GENERATE_KAFKA_PROJECTION_UPDATES: defines whether the real-time updater should send a message of update every time it writes the projection to Mongo. Default is false

Message schema

The message produced adheres to the following schema:

{
"type": "object",
"required": [
"projection",
"operationType",
"operationTimestamp",
"primaryKeys"
],
"properties": {
"projection": {
"type": "string",
"description": "The name of the projection"
},
"operationType": {
"type": "string",
"description": "The type of operation performed on the collection",
"enum": ["INSERT", "UPDATE", "DELETE"]
},
"operationTimestamp": {
"type": "date",
"description": "The timestamp at the operation execution"
}
"primaryKeys": {
"type": "array",
"items": {
"type": "string"
},
"description": "The primary keys of the projection"
},
"documentId": {
"type": "_id_",
"description": "The _id of the Mongo document"
}
"before": {
"type": "object",
"description": "The projection document as found on the database before the execution of the operation. `null` after an `INSERT`, `undefined` when not available."
},
"after": {
"type": "object",
"description": "The projection document as found on the database after the execution of the operation. `null` after a `DELETE`."
},
"__internal__kafkaInfo": {
"type": "object",
"description": "The __internal__kafkaInfo field of the document, if available."
}
}
}
info

The operationType describes the raw operation executed on Mongo, and can be different from the input operation type; for example an input UPSERT operation either results in a UPDATE message, or an INSERT message, depending on the existence of the document before the operation.
A virtual delete always results in an UPDATE, because the document still exists on Mongo after the execution with __STATE__: "DELETED", while an input real delete always results in a DELETE.

danger

For performance reasons, the projection state before of the projection is not returned after an update operation, because a costly find operation would be necessary to retrieve such information from the database. The only exception is an UPDATE resulting from a virtual delete: in this case the before is returned entirely.

For the same reason, the fields createdAt and documentId of the projections are returned only after an upsert inserting a new document or after a virtual delete.

Configuration

You need to create a configuration with the same path as the one you set in KAFKA_PROJECTION_UPDATES_FOLDER. Then, you have to create a configuration file kafkaProjectionUpdates.json inside that configuration. The json configuration should look like this one:

{
"MY_PROJECTION": {
"updatesTopic": "MY_UPDATES_TOPIC"
}
}

where

  • MY_PROJECTION is the name of the collection whose topic has received the message from the CDC.
  • MY_UPDATES_TOPIC is the topic where the updates message will be sent to.

Notice that you need to set a topic for each projection in the system. Different projection can share the same topic.

For example:

{
"registry-json": {
"updatesTopic": "registry-json.update"
},
"another-projection": {
"updatesTopic": "another-projection.update"
}
}

When the real time updater writes to Mongo in reaction to a CDC update, a message is sent to the related topic. For example, if a new projection is saved to the registry-json, an INSERT message is generated to the topic registry-json.update. The key of the resulting Kafka message will be the stringified JSON of the projection key value.

info

Notice that you can either set the topics for all the projections, or for a subset of them. So, for example, if you need to setup a Single View Patch operation, you may want to configure only the projections needed in such Single View.

Advanced topics

Primary Key update

Starting from v7.1.0, the Real-Time Updater supports the update of a Primary Key. In particular, it detects if received ingestion messages contain an updated Primary Key. In that case, the Real-Time updater automatically handles updating events on the existing records without requiring additional messages.

When the Real-Time Updater receives a Primary Key update, it triggers two different actions:

  1. the deletion of the old record with the old Primary Key
  2. the creation of a new record with the updated content and the updated Primary Key
note

In order to handle a Primary Key update, the Real-Time Updater needs to receive the following data:

  • a positional information about the current operation
  • the data values before the update
  • the data values after the update

More details about the configuration of these fields can be found inside each Kafka message format paragraph in this page.

danger

The Kafka message format based on the IBM InfoSphere Data Replication for DB2 CDC does not support the Primary Key update.

Snappy compression

Snappy is a compression and decompression library whose aim is to offer high speed data flow while still maintaining a reasonable compression ratio. Among the various types of compression supported by Kafka for its messages, there is also Snappy.

The main advantages of Snappy are:

  • Fast compression speed (around 250 MB/sec)
  • Moderate CPU usage
  • Stability and robustness to prevent crashing while still maintaining the same bitstream format among different versions
  • Free and open source

Note: For further information about Snappy, check the official GitHub page of the library.

Provided that the client's CDC (Change Data Capture) supports Snappy compression, the console is already predisposed for it.

caution

Snappy, like every other compression and decompression algorithm, will always increase the delay between production and consumption of the message, hence it is not advised for strong real-time relying applications; on the other hand it is well recommended for initial loads which tend to be a lot heavier.

Create ACL rules

It is possible to create ACL rules, you can set from user interface:

  • type: set topic type.
  • topic name: new or existent one.
  • pattern type: literal or prefixed. If you want to declare an ACL for each topic you should use literal.
  • operation: for each topic, you should set READ and WRITE operation.
  • permission: could be ALLOW or DENY. You should set ALLOW. Once created, by default permission are to deny all others operations.

Tracking the changes

From the v3.2.0 of the Real-Time Updater, inside the Projections and Projection Changes additional information about the Kafka message that triggered the Real-Time Updater are saved. This allows you to track the changes made as consequence of a Kafka message.

In particular, the following information are saved:

Projection changes

Into each element of the changes array of the projection change document are inserted the information about the message that triggered the projection change

  • topic: is the topic from which the Kafka message has been consumed
  • partition: is partition from which the Kafka message has been consumed
  • offset: is the offset of the message
  • key: is the key of the Kafka message
  • timestamp: is the timestamp of the message

Example:

{
"type": "sv_pointofsale",
"identifier": {
"ID_USER": "123",
},
"changes": [{
"state": "NEW",
"topic": "my-topic.development.my-system.my-projection.ingestion",
"partition": 0,
"timestamp": "2021-11-19T16:22:07.031Z",
"offset": "14",
"key": {
"ID_USER": "123",
},
}],
}

Projection

Into the projection is saved information about the last Kafka message that updated the projection. This information is saved inside a field named __internal__kafkaInfo in order to prevent collision with others projection fields.

The information saved are:

  • topic: is the topic from which the Kafka message has been consumed
  • partition: is partition from which the Kafka message has been consumed
  • offset: is the offset of the message
  • key: is the key of the Kafka message
  • timestamp: is the timestamp of the message

Example:

{
"ID_USER": "123",
"ADDRESS": "address_1",
"SURNAME": "ROSSI",
"TAX_CODE": "tax_code",
"__STATE__": "PUBLIC",
"createdAt": "2021-10-19T13:39:47.589Z",
"timestamp": "2021-11-19T16:22:07.031Z",
"updatedAt": "2021-11-19T16:22:07.052Z",
"__internal__kafkaInfo": {
"topic": "my-topic.development.my-system.my-projection.ingestion",
"partition": 0,
"timestamp": "2021-11-19T16:22:07.031Z",
"offset": "14",
"key": {
"ID_USER": "123",
},
}
}

Prevent projection to be overwritten

During a rebalancing or a massive initial load with multiple replicas of the real time updater, a batch of old messages that have not been committed yet could be read by the real time updater. In fact, Kafka ensures that messages are received, in order, at least once.

To prevent that old messages that have already updated the projection, overwrite the projection again, the environment variable FORCE_CHECK_ON_OFFSET is set by default to true.

This setting is strongly recommended when you have both insert/update and delete operations.

caution

At the moment this variable is set to true by default, but you can turn it off in order to adapt your services in case they need some fix.

Kafka group rebalancing behavior

If a Kafka group rebalancing happens after that a projection has already been updated, projection changes will be generated anyway and the Real Time updater will still try to commit though.

note

This behavior has been introduced from v4.0.0 and above. In previous versions instead, a rebalancing check was made after each operation, and when it happened, the service would stop without generating any projection change.

CA certs

Since service version 5.4.0, you can set your CA certs by providing a path to the certification file in the environment variable CA_CERT_PATH.

Single View Patch

info

This feature is supported from version 7.1.2 of the Real-Time Updater

As explained here, in order to arrange the Real-Time Updater for this specific operation, it's important to understand which are the projections impacted. Then, the following steps have to be followed:

  • Activate Projection Updates with the env GENERATE_KAFKA_PROJECTION_UPDATES
  • Configure the Projection Updates only for the specific Projection

Pause single topic's consumption on error

info

This feature is supported from version 7.1.4 of the Real-Time Updater

When the Real Time Updater encounters an error while consuming an ingestion message it will just pause the topic's consumption of that message, instead of crashing. This feature is disabled by default, to enable it use the PAUSE_TOPIC_CONSUMPTION_ON_ERROR environment variable.

To use this feature make sure you have properly configured your alerts in your grafana dashboards so you can quickly detect pauses on topics' consumption and act on the problematic message to resume the consumption. The ideal for that would be to set up some alerts based on the kafka_consumergroup_lag metric exposed by the kafka-exporter service configured in your project. More on configuring alarms on grafana here