Hi,
I am working with the 0.7.0-SNAPSHOT indeed.
This should do the same, I am testing it right now (if this does not leak, then
it comes from the camel integration, but these actions are the same done by the
camel component:
public static void main(String[] args) throws Exception {
try {
int i=0;
while(true){
//Thread.sleep(100); //IN MY EXAMPLE THE POLLING INTERVALL IS
100MS
i++;
PlcConnection connection = new
PlcDriverManager().getConnection("s7:tcp://192.168.178.10");
PlcReadRequest.Builder builder =
connection.readRequestBuilder();
builder.addItem("default","%DB1.DBW0:WORD");
PlcReadResponse response = builder.build().execute().get();
logger.info("{} Response Code: {} for
{}",i,response.getResponseCode("default"),response.getShort("default"));
connection.close();
}
} catch (Exception ex) {
ex.printStackTrace();
}
}
On 2020/02/27 16:37:49, Christofer Dutz <[email protected]> wrote:
> Hi Etienne,
>
> What version are you using? I know that Julian worked on quite some leaks and
> memory leaks were one the main reasons for the Christmas refactoring. But he
> also fixed a lot of things in the 0.6.0 branch. If you're already on
> 0.7.0-SNAPSHOT (develop branch) this is something we should definitely
> investigate.
>
> Would it be possible to provide a testcase where I could locally reproduce
> this behavior? Then I could let JProfiler loose on the example and see where
> the memory is leaking to.
>
> Chris
>
>
>
> Am 27.02.20, 17:06 schrieb "Etienne Robinet" <[email protected]>:
>
> Hello,
> after managing to do some read/write with camel, I tried to start
> fetching at specific interval. My goal was being able to fetch at regular
> pace (like every 500ms) to monitor some data. After running some tests, I
> came across an error:
>
> java.lang.OutOfMemoryError: Java heap space
>
> after rerunning some tests, I saw that the JVM will tak up to 4Gb before
> the crash. After check netstat in Powershell, I realized I was having 50k+
> different ports occupied on my localhost! I don't know if this has been
> lokked into yet or was never intended to be used like that, but after adding
> some Logs, I can tell that with my route, the app creates a Polling Consumer
> every tick that makes a plcCOnnection.close() when he's done. But after
> shutting down the context, a Plc4xConsumer class will also shutdown a
> connection. But still, in netsat I can see all these open connections for a
> bit of time. If anyone has an idea, that will be great! Here the route:
>
> <blueprint xmlns="http://www.osgi.org/xmlns/blueprint/v1.0.0";
> default-activation="lazy">
> <camelContext id="S7-PLC-Context"
> xmlns="http://camel.apache.org/schema/blueprint"; streamCache="true" >
> <route>
> <from uri="timer://foo?fixedRate=true&period=1000" />
> <pollEnrich>
>
> <constant>plc4x:s7:tcp://192.168.178.10?address=%25DB1.DBW0:WORD</constant>
> </pollEnrich>
> <setHeader headerName="fieldName">
> <constant>default</constant>
> </setHeader>
> <setHeader headerName="fieldQuery">
> <constant>%MW10:WORD</constant>
> </setHeader>
> <convertBodyTo type="java.lang.Short"/>
> <to uri="plc4x:s7:tcp://192.168.178.10" />
> <log message=" Response code:
> ${body.getResponseCode(default)} at ${date:now:HH:mm:ss}" loggingLevel="INFO"
> />
> </route>
> </camelContext>
> </blueprint>
>
>
> and the LOGGER :
> 17:04:55.883 INFO [pipe-bundle:start 67] Blueprint bundle
> s7-camel/0.1.0.SNAPSHOT has been started
> 17:04:55.884 INFO [Blueprint Event Dispatcher: 1] Attempting to start
> CamelContext: S7-PLC-Context
> 17:04:55.885 INFO [Blueprint Event Dispatcher: 1] Apache Camel 2.24.2
> (CamelContext: S7-PLC-Context) is starting
> 17:04:55.886 INFO [Blueprint Event Dispatcher: 1] StreamCaching is
> enabled on CamelContext: S7-PLC-Context
> 17:04:55.886 INFO [Blueprint Event Dispatcher: 1] JMX is enabled
> 17:04:55.912 INFO [Blueprint Event Dispatcher: 1] Instantiating new PLC
> Driver Manager with class loader BundleDelegatingClassLoader(s7-camel [67])
> 17:04:55.913 INFO [Blueprint Event Dispatcher: 1] Registering available
> drivers...
> 17:04:55.916 INFO [Blueprint Event Dispatcher: 1] Registering driver for
> Protocol s7 (Siemens S7 (Basic))
> 17:04:55.925 INFO [Blueprint Event Dispatcher: 1] StreamCaching in use
> with spool directory:
> C:\karaf\data\tmp\camel\camel-tmp-e0de8e4a-1313-44d0-a309-e211b894bd4d and
> rules: [Spool > 128K body size]
> 17:04:55.958 INFO [Blueprint Event Dispatcher: 1] Can read true can write
> true can subscribe true
> 17:04:56.018 INFO [Blueprint Event Dispatcher: 1] Route: route10 started
> and consuming from: timer://foo?fixedRate=true&period=1000
> 17:04:56.019 INFO [Blueprint Event Dispatcher: 1] Total 1 routes, of
> which 1 are started
> 17:04:56.020 INFO [Blueprint Event Dispatcher: 1] Apache Camel 2.24.2
> (CamelContext: S7-PLC-Context) started in 0.136 seconds
> 17:04:57.020 INFO [Camel (S7-PLC-Context) thread #17 - timer://foo]
> Instantiating new PLC Driver Manager with class loader
> BundleDelegatingClassLoader(s7-camel [67])
> 17:04:57.021 INFO [Camel (S7-PLC-Context) thread #17 - timer://foo]
> Registering available drivers...
> 17:04:57.021 INFO [Camel (S7-PLC-Context) thread #17 - timer://foo]
> Registering driver for Protocol s7 (Siemens S7 (Basic))
> 17:04:57.021 INFO [Camel (S7-PLC-Context) thread #17 - timer://foo]
> Creating POLLING CONSUMER
> 17:04:57.038 INFO [Camel (S7-PLC-Context) thread #17 - timer://foo] Can
> read true can write true can subscribe true
> 17:04:57.055 INFO [Camel (S7-PLC-Context) thread #17 - timer://foo]
> doStart POLLING Consumer
> 17:04:57.055 INFO [Camel (S7-PLC-Context) thread #17 - timer://foo]
> Creating read request
> 17:04:57.070 INFO [Camel (S7-PLC-Context) thread #17 - timer://foo]
> POLLING Consumer closed is false
> 17:04:57.070 INFO [nioEventLoopGroup-519-1] CHANNEL CLOSED : true
> 17:04:57.075 INFO [Camel (S7-PLC-Context) thread #17 - timer://foo] Value
> obtained java.lang.Short
> 17:04:57.077 INFO [Camel (S7-PLC-Context) thread #17 - timer://foo] Has
> optimizer : true
> 17:04:57.078 INFO [Camel (S7-PLC-Context) thread #17 - timer://foo]
> Writing data [0, 37, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
> 17:04:57.087 INFO [Camel (S7-PLC-Context) thread #17 - timer://foo]
> Response code: OK at 17:04:57
> 17:04:58.021 INFO [Camel (S7-PLC-Context) thread #17 - timer://foo]
> Creating POLLING CONSUMER
> 17:04:58.022 INFO [Camel (S7-PLC-Context) thread #17 - timer://foo] Can
> read true can write true can subscribe true
> 17:04:58.038 INFO [Camel (S7-PLC-Context) thread #17 - timer://foo]
> doStart POLLING Consumer
> 17:04:58.038 INFO [Camel (S7-PLC-Context) thread #17 - timer://foo]
> Creating read request
> 17:04:58.041 INFO [Camel (S7-PLC-Context) thread #17 - timer://foo]
> POLLING Consumer closed is false
> 17:04:58.041 INFO [nioEventLoopGroup-520-1] CHANNEL CLOSED : true
> 17:04:58.042 INFO [Camel (S7-PLC-Context) thread #17 - timer://foo] Value
> obtained java.lang.Short
> 17:04:58.042 INFO [Camel (S7-PLC-Context) thread #17 - timer://foo] Has
> optimizer : true
> 17:04:58.043 INFO [Camel (S7-PLC-Context) thread #17 - timer://foo]
> Writing data [0, 39, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
> 17:04:58.047 INFO [Camel (S7-PLC-Context) thread #17 - timer://foo]
> Response code: OK at 17:04:58
> 17:04:58.256 INFO [pipe-bundle:stop 67] Destroying container for
> blueprint bundle s7-camel/0.1.0.SNAPSHOT
> 17:04:58.257 INFO [Blueprint Event Dispatcher: 1] Stopping CamelContext:
> S7-PLC-Context
> 17:04:58.257 INFO [Blueprint Event Dispatcher: 1] Apache Camel 2.24.2
> (CamelContext: S7-PLC-Context) is shutting down
> 17:04:58.258 INFO [Blueprint Event Dispatcher: 1] Starting to graceful
> shutdown 1 routes (timeout 300 seconds)
> 17:04:58.259 INFO [Camel (S7-PLC-Context) thread #18 - ShutdownTask]
> Closed connection false
> 17:04:58.260 INFO [nioEventLoopGroup-518-1] CHANNEL CLOSED : true
> 17:04:58.260 INFO [Camel (S7-PLC-Context) thread #18 - ShutdownTask]
> Route: route10 shutdown complete, was consuming from:
> timer://foo?fixedRate=true&period=1000
> 17:04:58.260 INFO [Blueprint Event Dispatcher: 1] Graceful shutdown of 1
> routes completed in 0 seconds
>
>
>
> Etienne
>
>
>
>
