Hi DevP,
I suspect that it has to do with the logic in the HTTP_RESPONSE section. In your case, the iRule for the response fully relies on the var1 that's created in the HTTP_REQUEST section. As such, if HTTP_REQUEST already set the persistence record, there is no need to do it for the response as well. If the HTTP_REQUEST section did NOT create a var1 variable, the HTTP_RESPONSE section also won't do anything.
I would recommend removing the HTTP_RESPONSE section and then see if you still get error messages. If yes, please shout and we can dig a bit deeper. If so, would be great if you could move the log statement to the HTTP_REQUEST section as well.
For info; the reason why K7392 is mentioning the HTTP_RESPONSE section, is in the special case of cookies. In those requests, the initial request does NOT yet have the cookie and as such, the initial request does not get included in the persistence table, which means persistence for those sessions wouldn't start until the 2nd request, by which time it's already too late. As your requests should always have the "myuripart" in the request, you don't have that issue. - Thanks for referencing the K article though, it helps better understand the code 😉
Hope this helps.