returns 32

2016-04-01 16:47:35 NativeLoggerBridge [INFO] CONMGR: Details: writeData(); None; 0; -808; getSockOpt() failed on fd 85 returns 32;2016-04-01 16:47:35 NativeLoggerBridge [INFO] Connector: cycleNextAddress, Cycling from lawin:11101 to agila:11101 because connection was dropped

The error code returns 32 is translated as broken pipe. This means that the Gateway and Web Dashboard got disconnected. Check if the Gateway is online.

returns 104

2016-03-29 12:48:49 NativeLoggerBridge [INFO] CONMGR: Details: readData(); None; 0; -808; getSockOpt() failed on fd 88 returns 104

The error code returns 104 is translated as connection reset by peer. This may mean 2 things:

1. The Web Dashboard is overloaded.
2. The Gateways it is connected to are busy.

Addressing #1

• Option #1: Install a 64-bit Web Dashboard.
• Option #2: Split the load by defining other dashboards to another Web Dashboard.

Monitor also the Web Dashboard's performance by enabling its JMX port and configuring a JMX sampler to extract the Web Dashboard's performance MBeans.

Addressing #2

Monitor the affected Gateway's load by performing a 24-hour ad-hoc load monitoring collection (This can be changed to a 1-hour or 30-minute collection depending on the busyness of the Gateway). Once done, the Gateway produces a stats.xml file. Analyze the contents of the stats.xml file and apply the needed tuning/changes. The stats.xml can also be sent to ITRS Support for analysis along with the Gateway diagnostics file and Gateway log file by raising a support ticket.