Sequence number

26 views
Skip to first unread message

Bjarne Johansson

unread,
Mar 18, 2024, 4:17:47 PMMar 18
to open62541
Hi everyone,
This is a very specific question, concerning UADP and PubSub. 
It seems like the sequence number in the GroupHeader and DataSetMessageHeader are ignored by the subscriber. Is that intended even if the content mask are set to include sequence numbers? 

Thanks and regards, 
Bjarne 

Julius Pfrommer

unread,
Mar 18, 2024, 6:50:26 PMMar 18
to Bjarne Johansson, open62541
Yes, you are right.
Due to the (possibly unreliable) transport, PubSub messages might get lost or reordered during transport.
The SequenceNumber is defined in the standard to filter out old messages after an unintentional reorder.

Since there can be several Publishers received by a ReaderGroup the SequenceNumber needs to be tracked *per Publisher*.
We currently don't track the individual Publishers on the receiving end.

If you need this handled correctly, please open an issue on Github.

Regards, Julius

--
You received this message because you are subscribed to the Google Groups "open62541" group.
To unsubscribe from this group and stop receiving emails from it, send an email to open62541+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/open62541/75600055-28ba-402a-a804-34cad21465acn%40googlegroups.com.

Gunter Pfeiffer

unread,
Mar 19, 2024, 2:24:37 AMMar 19
to open62541
Maybe the root-cause is because of this problem:
Setting UA_UadpDataSetMessageContentMask has no impact, default configuration is used · Issue #6334 · open62541/open62541 · GitHub

Because the  UadpDataSetMessageContentMask ist not respected, so you can't configure any fields in the headerlayout of the data message... 

Bjarne Johansson

unread,
Mar 19, 2024, 4:58:23 AMMar 19
to open62541
Thank you both for the replies. I just wanted to confirm what I observed.
I'm using the UA_PUBSUB_RT_FIXED_SIZE WriterGroup config. For that #6334 is not a problem. 
I will most likely fork for the prototyping I'm doing, in that fork I might add the sequence number check. 

Regards
Bjarne

Reply all
Reply to author
Forward
0 new messages