Single http receiver channel with timeout issue stops whole message traffic

[Former Member]

Hi PI/PO experts,

we faced a strange behaviour on our PI/PO 7.31 SP8 Java only installation:

An external partner changed the URL of his http server and our http receiver channel ran into timeout issues as we sent a few IDoc from our ECC System via PI. At the same time we sent thousands of other IDocs via PI and http Adapter to another external Partner as well as a few hundert of other messages with other adapter types. All these messages ran into status "Scheduled" and the whole message traffic literally stopped! The Server itself runs flawless, in terms of CPU health, access to development tools, monitoring etc. during that time.

As we stopped the channel with the wrong URL the related messages ran into status "Error" and the other messages started over an were sent successfully and set in status "Delivered".

We opened a support ticket, explored this issue on scn and SAP Support colleagues told us to increase the FCAWorkerThreads and from a scn discussion we changed the values for maxParallelQueues and QueueTypes. But nothing helped, as we reproduced this issue over oand over again on our QUAL environment (setting FCAWorkerThread to 50, 20, 30; Setting maxParallelQueues to 5 and 2, seeting  QueueTypes to "Recv, IcoAll", testing with different amount of IDcos we sent via PI). Every time we let messages ran into timeout issues, all other messages were stuck in Status "Delivered".

Did you face something similiar and what did you do? Any help is welcome!

Thanks

Daniel

P.S.: in the specific Receiver http channel no special parameters were set, no modules added and so on. Just target URL, path and access data.

[bhavesh_kantilal]

Hello Daniel,

Are you using a HTTP adapter or a HTTP_AAE?

Regards,

Bhavesh

[Former Member]

Hi Bhavesh,

it´s HTTP_AAE.

Regards

Daniel