ICM_HTTP_CONNECTION_FAILED for a wrong Host Name

[Former Member]

Hi Experts,

I have a strange and urgent query here. Yesterday Basis has added a virtual host to the Production PI box to the already existing server nodes, but the virtual host was not correctly created thus messages in XI which were sent by ICM to this host started failing. Once we started getting the errors and after the initial findings, we got the new host shut down , but the problem is the messages which were already failed in SXI_MONITOR with the error

"Error while receiving by HTTP (error code: 400,error text: ICM_HTTP_CONNECTION_FAILED)" becaceuse of the faulty host name are still failing with the same reason.

We have already bounced the PI server and did full cache refresh twice, but whenever we are restarting the message in moni we are getting the same error.;

Is there any way the message can be restarted from the initial point, or any way that the faulty host name can be removed from header data of the failed message.

A quick reply will be highly appreciated.

Thanks alot.

Regards,

Rahul

[Former Member · Accepted solution]

Hi All,

Thanks for the help, but as I thought there would not be any feasible option to restart these messages in Production because of:

1, No authorizations to edit the messages

2. Very high number of messages,so can not use RWB to resend

3. Creating a report to resend all the messages, many of which had very big payloads, is very time consuming.

Thus we stick to the basic and tried another full cache refresh and a complete bounce of server, this cleared the erroneous entries of the stopped host from cache and the messages were reprocessed swiftly this time.

Thanks again for all the help and support.

Regards,

Rahul

[ambrish_mishra]

Rahul,

Curious to know if the cache update/system restart removed Host entry from the messages in error and how the message could be reprocessed with the stopped host in the SOAP header.

Details would be helpful since the message is marked Answered.

Ambrish

[former_member201264]

Hi Rahul,

Consider the following points:

1. If you do not have any other option to resolve your issue and

2. If you have authorization to send the message from RWB.

3. If you have little number of messges to re-process.

4. Discusss with tech team and functional team to send from RWB.

Then get the XML Paylod from from the error message from SXMB_MONI and then you can proceed to send the data from Integration Engine of RWB.

This is one of my opinion and not the solution.

Ask your manager and get approval emails from them and then proceed.

Regards,

Sreeni.

[former_member184720]

Hi Rahul -

I don't think we have an option to restart a message from initial point but won't you be able to ask the source system to resend?

may be based on the time frame you can request them to send the messages again to PI.

If there are lots of messages& not possible to resend from source then may be you can talk to ABAP expert to write a report to download the XML's using the message ID's. Later just change the sender adapter to File and process the XML's again- I don't think you do this

Regards,

Hareesh

[ambrish_mishra]

Hi Rahul,

Is there any way message can be restarted from initial point, the answer is no since the faulty host name is part of the message header.

You can try editing the message(if you have authorization) and change the Runtime parameter for the failed message under Inbound Message->SOAP Header for host and then restart the message, it may work.

I have never faced this so I am not 100% sure.

Ambrish