For partners that build or integrate commercially available applications and service solutions with the Oracle Cloud Platform
For partners that provide implementation or managed services around Oracle Cloud Applications
Hi,
Could anybody throw some light what exactly happens in the background when an OIC agent is installed ,how exactly it communicates with OIC instance and how does the agent ensures high availability.
Thanks
Hi Ritu,
On a high level Connectivity Agent will poll OIC agent channels for messages. When ever it found messages it will pick and process and sends the response.
Agent high availability is in works. Will be available in future release.
Adding to what Nagi said:
Connectivity Agent installed behind the firewall makes an outbound https request to OIC / ICS for all communications. This communication is protected using TLS ( transport level security). And there’s absolutely no need for customer to open up the inbound port on the firewall
ravi
Hi Nagi and Ravi,
" Connectivity Agent will poll OIC agent channels for messages." Does this means that OIC agent no more listens to OMCS?
Thanks
Correct! There is no OMCS involved in OIC Connectivity Agent.
Hi Experts,
Thanks for the quick updates.What I am trying to understand is how OIC agent is communicating with OIC instance.When we say that agent polls the channels, what channels are we talking about here?I was going through the agent properties file as below:
icsPlatform=ExternalCompute
proxyAuth=
adminProfile=
authenticatedProxy=false
agentTransport=aq
proxy_nonProxyHosts=
namespace=CONNECTIVI_AGENT_GROUP
omcsUri=
proxy_host=
agent_port=
managedserverport=
agentWorkerThreads=40
proxy_port=
externaljars=
onPremiseAgentVersion=18.2.5.0.0
agentInstanceId=CA-SAMPLE-XXXXXX
icsRESTBaseURI=https\://***.oraclecloud.com\:443
Does it mean this channel is AQ channel(on OIC instance) where the agent is polling?Are there any oracle documentation which talks about agent communication architecture as a whole.
From HA availability perspective, lets say for example, one agent(agent 1 & agent 2) is installed on each of the on premise managed node(node 1 & node 2).If one of the agent goes down then how is the load balancing and high availability managed here?
Thanks..
Ritu,
Connectivity Agent communicates with OIC / ICS over standard HTTP protocol.
Here are the salient features of the Connectivity Agent that are vividly described in the documentation.
No ports are opened on the on-premises system for communication.
All communication is secured using SSL.
The on-premises connectivity agent registers with Oracle Autonomous Integration Cloud over SSL using the provided Oracle Autonomous Integration Cloud credentials.
The on-premises connectivity agent checks for work by making outbound requests through the firewall.
The on-premises connectivity agent can use a proxy to access the internet (the same proxy as other internal applications and browsers use). Authentication support for outbound proxy access is provided.
The on-premises connectivity agent connections are configured by the agent retrieving the configuration details from Oracle Autonomous Integration Cloud.
The on-premises connectivity agent processes requests by pulling messages from Oracle Autonomous Integration Cloud across SSL.
The on-premises connectivity agent posts responses by pushing messages to Oracle Autonomous Integration Cloud across SSL.
All communication is initiated by the on-premises connectivity agent.
No private SOAP-based web services are exposed.
No existing J2EE container is required to deploy the on-premises connectivity agent.
No data is persisted in the on-premises agent.
Regarding HA, when the feature becomes available, the 2 agents that are associated to a single agent group form a virtual cluster. As you are aware, Agent Group is specified during agent installation. All of the work requests emanating from integration flows having connections bound to this agent group will be available for either of the agent to process. It is the 'Competing Consumer Pattern' that is in the play.
Ravi
Ritu,
Connectivity Agent communicates with OIC / ICS over standard HTTP protocol.
Here are the salient features of the Connectivity Agent that are vividly described in the documentation.
No ports are opened on the on-premises system for communication.
All communication is secured using SSL.
The on-premises connectivity agent registers with Oracle Autonomous Integration Cloud over SSL using the provided Oracle Autonomous Integration Cloud credentials.
The on-premises connectivity agent checks for work by making outbound requests through the firewall.
The on-premises connectivity agent can use a proxy to access the internet (the same proxy as other internal applications and browsers use). Authentication support for outbound proxy access is provided.
The on-premises connectivity agent connections are configured by the agent retrieving the configuration details from Oracle Autonomous Integration Cloud.
The on-premises connectivity agent processes requests by pulling messages from Oracle Autonomous Integration Cloud across SSL.
The on-premises connectivity agent posts responses by pushing messages to Oracle Autonomous Integration Cloud across SSL.
All communication is initiated by the on-premises connectivity agent.
No private SOAP-based web services are exposed.
No existing J2EE container is required to deploy the on-premises connectivity agent.
No data is persisted in the on-premises agent.
Regarding HA, when the feature becomes available, the 2 agents that are associated to a single agent group form a virtual cluster. As you are aware, Agent Group is specified during agent installation. All of the work requests emanating from integration flows having connections bound to this agent group will be available for either of the agent to process. It is the 'Competing Consumer Pattern' that is in the play.
Ravi