G
Gareth Vincent
Guest
We are often experiencing problems on our Solaris server with OE 11.5.1 whereby agents are getting locked in "SENDING" state. These agents often take up 5 - 10% of CPU utilization so we are constantly having to monitor and kill these runaway processes. So far I've identified that this is typically caused by bad connections from the end users. The users get disconnected yet the appserver remains in "SENDING" state. We have other customers running on Centos with the same version of OE and same source base yet they are not experiencing the same problem. The difference is that these servers are hosted on premise so line issues are not really a factor as they are connecting over the LAN. We have not ruled out coding issues but I would like to know if there is anything from a config point of view that could minimize or eliminate these "hanging" agents.
Continue reading...
Continue reading...
