OCPP server fail to send commands to the charging stations when two connections are established

[katsuya]

Description

When two connections are established to the same charge point A, the following bug occurs, causing OpenEMS to fail in properly handling the previous connection:

• When connection B is established, and a lostSession is triggered in activeEvcSessions for connection A, it incorrectly affects charge point A as a whole. This results in connection B malfunctioning as well.

The following issues should be addressed (resolving the first issue alone might be sufficient):

1. When a new connection is established, the old connection must be correctly removed from activeEvcSessions.

   • openems/io.openems.edge.evcs.ocpp.server/src/io/openems/edge/evcs/ocpp/server/MyJsonServer.java

     Line 105 in 263377b

     }

2. When the old connection is disconnected, it should ensure that the lostSession for the charge point referenced by the new connection is not invoked.

   • openems/io.openems.edge.evcs.ocpp.server/src/io/openems/edge/evcs/ocpp/server/MyJsonServer.java

     Line 117 in 263377b

     ocppEvcs.lostSession();

3. Disconnection of the old connection should not remove the charge point from ocppSessions associated with the new connection.

   • openems/io.openems.edge.evcs.ocpp.server/src/io/openems/edge/evcs/ocpp/server/MyJsonServer.java

     Line 128 in 263377b

     MyJsonServer.this.parent.ocppSessions.remove(ocppId);

Screenshots

No response

Operating System

Ubuntu

How to reproduce the Error?

1. Connect charge point A to the OCPP server.
2. Establish another connection between charge point A and the OCPP server without disconnecting the first connection.
3. Disconnect the first connection.

[sfeilmeier]

@sebastianasen: FYI. Any clue?
@katsuya: OCPP is not actually used right now on our side, so this has no high priority ATM.

[katsuya]

@katsuya: OCPP is not actually used right now on our side, so this has no high priority ATM.

@sfeilmeier That's really unfortunate. We've started using it, so if we send a patch from our side, would you consider reviewing it?

[Sn0w3y]

Hello @katsuya see my PR :)

Please tell us if it works for you !

[Sn0w3y]

does it work @katsuya ?

[sfeilmeier]

@sfeilmeier That's really unfortunate. We've started using it, so if we send a patch from our side, would you consider reviewing it?

Sure. I just wanted to point out that we are currently not working on this issue. If there is a working and reviewed patch, I'll of course merge it.

[Sn0w3y]

Hello @katsuya could you already test my PRs ?

[katsuya]

I previously reported that there might be some odd behavior related to multithreading (#2908 (comment)). However, after testing over the weekend in an environment where the Wi-Fi connection was intermittently dropping, we confirmed that there are no other issues aside from the two we have already reported (this issue #2906 and #2909). All other problems were related to the OCPP client side.

After testing the patch this weekend, I have a few concerns and will comment on them individually. @Sn0w3y

[Sn0w3y]

I previously reported that there might be some odd behavior related to multithreading (#2908 (comment)). However, after testing over the weekend in an environment where the Wi-Fi connection was intermittently dropping, we confirmed that there are no other issues aside from the two we have already reported (this issue #2906 and #2909). All other problems were related to the OCPP client side.

After testing the patch this weekend, I have a few concerns and will comment on them individually. @Sn0w3y

Yes but my PRs should fix the 2 Issues :D did they?

[katsuya]

@Sn0w3y I left a comment on the PR (#2908 (review)). Could you take a look when you have time?