SocketXP Agent Change Log

Version 2.0.0

From version 2.0.0 onwards, SocketXP agent will connect to TCP port 9443 on the gateway. Older version agents (1.4.5 and earlier versions) will continue to connect to TCP port 80 on the gateway. The gateway will continue to support both the legacy and newer agent versions.
Create a single (reverse proxy) tunnel from the device to the gateway to remotely access any number of services(SSH, HTTP web server, VNC, RDP, Database etc.) running in the device. Older agent versions(1.4.5 and earlier), create one tunnel per service that needs to be remotely accessed. The new agent will create just one tunnel per device irrespective of the number of services in the device that needs remote access.

Bring back the exponential back-off mechanism to retry temporary failures in connecting with the gateway.
Increased the Linux systemd service RestartSec interval setting in the /etc/systemd/system/socketxp.service config file to 60 secs from 5 secs. These changes are extremely important to avoid bombarding the gateway during init after a server restart.

Under certain circumstances, SocketXP agent's TCP(TLS) connection to the gateway may show up as active indefinitely, eventhough the gateway has rebooted. SocketXP agent may not timeout and retry establishing a new connection to the gateway. The fix is to prevent disabling the OS's TCP protocol keepalive messages prematurely before starting the heartbeat/ping messages. This bug fix applies to agent version 1.4.0 and above.

When running SocketXP agent in the background in IoT Slave Mode, exit the agent on failure to reconnect with the SocketXP Gateway. Systemd will respawn the agent.

Fixed a bug in the SocketXP agent running in IoT Slave mode that uses a config file based configuration. The bug made the IoT Slave mode agent to connect to a single IoT device despite having multiple IoT Slave mode tunnel configs in the config.json file. This bug was introduced in 1.4.1.
Added support to stop and exit SocketXP agent, instead of retrying in a loop, when the network interface used by the agent to reach internet goes down. This may take upto 10 secs. The agent will be restarted by systemd and may try reaching internet via an alternate network interface or keep waiting until the internet connection come up.
Exit from the SocketXP agent when the agent detects ping (or keepalive) failure on the control channel. Linux systemd will restart the agent after a few secs and retry connecting with the SocketXP Cloud Gateway. Previously, the agent used to retry the connection to the gateway in a loop.

Added support for optionally turning on Asset Tracking feature. By default this feature is disabled.
Fixed a bug that prevented log file from being created in "/var/lib/socketxp" when "--config" option is used.

Added support for File Monitoring. It provides users the option to turn ON file monitoring and receive webhook alerts whenever a log file or config file in the device get modified.
Restart the agent whenever a new monitoring config is downloaded from the gateway.
Stopped TCP protocol stack from sending keepalalives every 15 seconds, which overrides the ping_interval(default 90 seconds) of SocketXP agent.
Log file rotation.

Added support for Device Resource(CPU, Memory, Disk) Monitoring. It provides users the option to turn ON device resource monitoring and receive webhook alerts when the device resource utilization exceeds a configured threshold value. By default this feature is disabled.
Provided an option to change the default working directory of the SocketXP agent using the "--work-dir" flag.