We are doing bulk REST operation, that lasts for more than 15 mins , we are using wildfly undertow load balancer. We see that load balancer is retriggering the rest call internally after 11 mins even though original REST call have not completed yet. could anyone please suggest where we are going wrong? We have frontend : undertow load balancer and backend : 3 nodes of widlfy we are using CPU metric for load sharing.
When we triggered bulk update from front-end , we observed below logs:
Initial request went to IVM-0 (i.e. this is wildlfy node)
2019-02-28 14:36:41,849 [default task-9] INFO c.n.w.r.f.CustomFilter:doFilter:57 - Rest URL=<SI /myapp/rest/someServie/update>
Second request went to IVM-10 (i.e. this is wildlfy node)
2019-02-28 14:47:53,871 [default task-6] INFO c.n.w.r.f.CustomFilter:doFilter:57 - Rest URL=<SI /myapp/rest/someServie/update>
Initial request completed by (IVM-0) :
2019-02-28 14:52:27,935 [default task-9] INFO c.n.w.o.SomeOperations:update:1175 - LOCK released
access logs
10.247.84.107 - - [[28/Feb/2019:14:52:27 +0530]] "PUT /myapp/rest/someServie/update HTTP/1.1" 200 29 946128
Third request went to IVM-10 :
2019-02-28 14:59:05,898 [default task-3] INFO c.n.w.r.f.CustomFilter:doFilter:57 - Rest URL=<SI /myapp/rest/someServie/update>
Failed as already one request going on
access logs
10.137.84.207 - - [[28/Feb/2019:14:59:07 +0530]] "PUT /myapp/rest/someServie/update HTTP/1.1" 575 71 2026 ( Bulk update in progress shown in front-end)
Second request completed one IVM-10
2019-02-28 15:01:17,361 [default task-6] INFO c.n.w.o.SomeOperations:update:1175 - LOCK released
Note: We observed this issue only with Mozilla Firefox browser