SAP BAS: exactly every other request fails
source link: https://answers.sap.com/questions/13725599/sap-bas-exactly-every-other-request-fails.html
Go to the source link to view the article. You can view the picture content, updated content and better typesetting reading experience. If the link is broken, please click the button below to view the snapshot at that time.
I am developing for several months now in BAS, today something stranges happens to me, working on my fiori app.
Exactly every other request to the on premise SAP ABAP system fails with a 500 Server error. Also they don't show up in SAP Cloud connector (the failing ones, that is). The requests in between work perfectly fine.
For example the requests to the oData metadata, that look like this: https://port8080-workspaces-ws-q8mdc.us10.trial.applicationstudio.cloud.sap/sap/opu/odata/sap/ZNAME_OF_MY_CDS_SERVICE/$metadata?sap-client=001&sap-language=EN
Every secord request works, the ones in between fail. I am confused, as with several years of development experience, I have never seen something like this.
And it's only the requests to the on premise system, that show this behaviour, like there would be some round robin, that uses a wrong route every other requests.
Another debugging tracing the raw traffic back to the right port for a successfull request gives 6 incoming packets:
09:18:58.748295 IP ec2-44-198-121-126.compute-1.amazonaws.com.https > 10.0.0.36.58368: Flags [.], ack 127, win 570, options [nop,nop,TS val 4173027627 ecr 1597118441], length 0
09:18:58.748512 IP ec2-44-198-121-126.compute-1.amazonaws.com.https > 10.0.0.36.58368: Flags [.], ack 166, win 570, options [nop,nop,TS val 4173027627 ecr 1597118441], length 0
09:18:58.748530 IP ec2-44-198-121-126.compute-1.amazonaws.com.https > 10.0.0.36.58368: Flags [P.], seq 0:39, ack 166, win 570, options [nop,nop,TS val 4173027627 ecr 1597118441], length 39
09:18:59.106140 IP ec2-44-198-121-126.compute-1.amazonaws.com.https > 10.0.0.36.58368: Flags [.], seq 39:1427, ack 166, win 570, options [nop,nop,TS val 4173027984 ecr 1597118543], length 1388
09:18:59.106602 IP ec2-44-198-121-126.compute-1.amazonaws.com.https > 10.0.0.36.58368: Flags [P.], seq 1427:1564, ack 166, win 570, options [nop,nop,TS val 4173027984 ecr 1597118543], length 137
09:18:59.251741 IP ec2-44-198-121-126.compute-1.amazonaws.com.https > 10.0.0.36.58368: Flags [.], ack 201, win 570, options [nop,nop,TS val 4173028130 ecr 1597118904], length 0
while for a failing request it's just two:
09:19:02.957304 IP ec2-44-198-121-126.compute-1.amazonaws.com.https > 10.0.0.36.58368: Flags [.], ack 328, win 570, options [nop,nop,TS val 4173031836 ecr 1597122650], length 0
09:19:03.083612 IP ec2-44-198-121-126.compute-1.amazonaws.com.https > 10.0.0.36.58368: Flags [P.], seq 1564:1699, ack 328, win 570, options [nop,nop,TS val 4173031962 ecr 1597122650], length 135
I am using the latest version of SAP Cloud Connector 2.14.2 on Linux, and as stated in the very first post, everything was working fine until friday.
To rule out, that there is something weired with my machine, I now set up another Cloud Connector on Win10 and I'm using that one, seing the same behaviour.
The 'every other request failing' is even valid for the same request from different browsers on the same computer, the same request from different computers or different requests to urls from different computers.
Another thing I tried is deleting my BAS instance, creating a new one, just git pulling in the code, npm installing and npm start. Same behaviour.
I also went back a day, to use the code base I was developing on for more than 8 hours without problems. No change.
Multiple back end application server nodes where the ICM is not functioning on one of them maybe?
Recommend
About Joyk
Aggregate valuable and interesting links.
Joyk means Joy of geeK