Hi Everyone,
IFS Cloud 23R2
We are working towards getting load balancing between 3 middletier servers working for our IFS cloud installation. This works ok for the most part, however we are noticing that timeouts are occurring right around one minute. In the application when this occurs there is no error message; whatever you were doing just stops so it is unclear sometimes that any timeout occurred.
Examples:
- Do a search on a screen that takes longer than a minute to return: Just see “No Results” even though results should have been returned
- Order a report that takes longer than a minute.
After clicking finish to order the report, the loading icon spins for one minute, and then the buttons became available again as if nothing was pressed and nothing happened. (The ordered report still shows up in the report archive once it is complete however)
Reviewing the logs on one of the middle tier servers here are some standouts that seem to indicate the issue:
| 2024-07-25T15:31:29.033251728Z stderr F 2024/07/25 15:31:29 5info] 50#50: *926747 client 10.1.136.206 closed keepalive connection |
| 2024-07-25T15:31:30.523125387Z stderr F 2024/07/25 15:31:30 5warn] 50#50: *926750 *lua] openidc.lua:1456: authenticate(): using deprecated option `opts.redirect_uri_path`; switch to using an absolute URI and `opts.redirect_uri` instead, client: 10.1.136.206, server: <REDACTED>, |
| 2024-07-25T15:32:36.177829765Z stdout F {"message": "POST /main/ifsapplications/projection/v1/SuppAgeAnalysisRep.svc/VirtualOrderReports(Objkey='1E122FC96C491CBDE06375C010AC0863')/IfsApp.SuppAgeAnalysisRep.VirtualOrderReport_OrderReport HTTP/1.1", "time": 1721921556.177, "http.response.body.bytes": 0, "host": "<REDACTED>", "address": "<REDACTED>", "request_length": 8194, "http.request.method": "POST", "uri": "/main/ifsapplications/projection/v1/SuppAgeAnalysisRep.svc/VirtualOrderReports(Objkey='1E122FC96C491CBDE06375C010AC0863')/IfsApp.SuppAgeAnalysisRep.VirtualOrderReport_OrderReport", "http.response.status_code": 499, "user_agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/126.0.0.0 Safari/537.36", "resp_time": 65.659, "upstream_addr": "10.152.183.178:8080" |
| 2024-07-25T15:32:36.17785982Z stderr F 2024/07/25 15:32:36 1info] 50#50: *926750 epoll_wait() reported that client prematurely closed connection, so upstream connection is closed too while sending request to upstream, client: 10.1.136.206, server: <REDACTED>, |
the IP address 10.1.136.206 exists in a subnet on the middle tier server.
Our infrastructure specialist thinks that might related to cluster management.
Has anyone else come across this issue?
Thanks,
LEdwards