Hi Kai,
that document is exactly the one I'm currently using.
What I'm currently working on is updating the current plugin to not schedule
and handle the connection stuff manually, but use the scraper component of
PLC4X.
Also is the current configuration not production ready and I'll be working on
to make it more easily usable.
But it will definitely not hurt to have some Kafka Pro have a look at what we
did and propose improvements. After all we want the thing to be rock-solid :-)
Chris
Am 31.07.19, 17:03 schrieb "Kai Wähner" <megachu...@gmail.com>:
Hi Chris,
great that you will work on the connector.
I am not deep technical, but if you need guidance from Kafka Connect
experts, I can connect you to a Confluent colleague to can help with best
practices for building the connector.
For example, we have implemented a wildcard option into our MQTT Connector
to map MQTT Topics to Kafka Topics in a more flexible way (e.g. 1000s of
cars from different MQTT Topics can be routed into 1 Kafka Topic). This
might also be interesting for this connector as you expect to various PLCs.
This guide might also help:
https://www.confluent.io/wp-content/uploads/Verification-Guide-Confluent-Platform-Connectors-Integrations.pdf
On Wed, Jul 31, 2019 at 4:39 PM Christofer Dutz <christofer.d...@c-ware.de>
wrote:
> Hi all,
>
> I am currently planning on cleaning up the Kafka Connect adapter a little
> as this was implemented as part of a proof of concept and is still I a
> state I wouldn’t use in production ;-)
> But a lot has happened since then and I’m planning on making it a really
> usable tool in the next few days.
>
> A lot has changed since we created the integration module QT3 2018 and I
> would like to refactor it to use the Scraper for the heavy lifting.
>
> Currently a user has to provide a parameter “query” which contains a
> comma-separated list of connection-strings with appended address. This is
> purely unmanageable.
>
> I would like to make it configurable via JSON or Yaml file.
>
> I think it would make sense to define groups of fields that are collected
> on one device at an equal rate. So it’s pretty similar to the scraper
> example, however I would like to not specify the source in the job, but
the
> other way around.
> When specifying the “sources” I would also provide which jobs should run
> on a given collection.
> As the connector was initially showcased in a scenario where data had to
> be collected on a big number of PLCs with equal specs,
> I think this is the probably most important use-case and in this it is
> also probably more common to add new devices to collect standard data on
> than the other way around.
>
> Also should we provide the means to also set per connection to which
> kafka-topic the data should be sent to.
> We could provide the means to set a default and make it optional however.
> When posting to a topic we also need to provide means for partitioning, so
> I would provide sources with an optional “name”.
> Each message would not only have the data requested, but also the
> source-url, source-name and the job-name with a timestamp.
>
> So I guess it would look something like this:
>
> #
>
----------------------------------------------------------------------------
> # Licensed to the Apache Software Foundation (ASF) under one
> # or more contributor license agreements. See the NOTICE file
> # distributed with this work for additional information
> # regarding copyright ownership. The ASF licenses this file
> # to you under the Apache License, Version 2.0 (the
> # "License"); you may not use this file except in compliance
> # with the License. You may obtain a copy of the License at
> #
> # http://www.apache.org/licenses/LICENSE-2.0
> #
> # Unless required by applicable law or agreed to in writing,
> # software distributed under the License is distributed on an
> # "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
> # KIND, either express or implied. See the License for the
> # specific language governing permissions and limitations
> # under the License.
> #
>
----------------------------------------------------------------------------
> ---
> # Defaults used throughout all collections
> defaults:
> # If not specified, all data goes to this topic (optional)
> default-topic: some/default
>
> # Defines connections to PLCs
> sources:
> # Connection to a S7 device
> - name: machineA
> # PLC4X connection URL
> url: s7://1.2.3.4/1/1
> jobs:
> # Just references the job "s7-dashboard". All data will be published
> to the default topic
> - name: s7-dashboard
> # References the job "s7-heartbeat", however is configures the
> output to go to the topic "heartbeat"
> - name: s7-heartbeat
> topic: heartbeat
>
> # Connection to a second S7 device
> - name: machineB
> url: s7://10.20.30.40/1/1
> # Sets the default topic for this connection. All jobs data will go to
> "heartbeat"
> topic: heartbeat
> jobs:
> - s7-heartbeat
>
> # Connection to a Beckhoff device
> - name: machineC
> url: ads://1.2.3.4.5.6
> topic: heartbeat
> jobs:
> - ads-heartbeat
>
> # Defines what should be collected how often
> jobs:
> # Defines a job to collect a set of fields on s7 devices every 500ms
> - name: s7-dashboard
> scrapeRate: 500
> fields:
> # The key will be used in the Kafka message to identify this field,
> the value here contains the PLC4X address
> inputPreasure: %DB.DB1.4:INT
> outputPreasure: %Q1:BYTE
> temperature: %I3:INT
>
> # Defines a second job to collect a set of fields on s7 devices every
> 1000ms
> - name: s7-heartbeat
> scrapeRate: 1000
> fields:
> active: %I0.2:BOOL
>
> # Defines a third job that collects data on Beckhoff devices
> - name: ads-heartbeat
> scrapeRate: 1000
> fields:
> active: Main.running
>
> I think it should be self-explanatory with my comments inline.
>
> What do you think?
>
> Chris
>
>
