When you have several applications or when you have multiple tiers in an application, creating applications using the public APIs or the user interface becomes a long process. VMware Aria Operations for Networks auto-discovers applications and tiers and enables you to add them automatically, which reduces a lot of manual effort.
In VMware Aria Operations for Networks, you can discover applications based on:
- Tags (VMware vCenter Server or AWS tags)
- VM Names
- ServiceNow
- Flows
- Advanced properties such as a combination of VM names, VM tags, NSX-V security tags, and security groups.
Note: In VMware Aria Operations for Networks, NSX tags are displayed as VM tags. Therefore, by default, NSX tags are also supported in application discovery.
An example of the application discovery construct
- Added VMware vCenter server as a data source.
- Added four VMs in your data center - VM1, VM2, VM3, and VM4.
- Defined tags (key-value) that defines the application names to which each VMs belong.
- Defined tags (key-value) that defines the tier to which each VMs belong.
VM Name | Key-value tags |
---|---|
VM1 |
|
VM2 |
|
VM3 |
|
VM4 |
|
Discovering Applications based on Tags
VMware Aria Operations for Networks, you can define a grouping criteria for application discovery for these tags.
Application | Tier | VM |
---|---|---|
MyApplication1 | App | VM1 |
MyApplication1 | Web | VM2 |
MyApplication2 | App | VM3 |
MyApplication2 | Web | VM4 |
Discovering Applications and Tiers based on VM Names
If the VM names are defined in the following format:
ApplicationName : Tier : VMName
.
MyApplication1 : App : VM1 MyApplication1 : Web : VM2 MyApplication2 : App : VM3 MyApplication2 : Web : VM4
- App Regex:
(.*)_(.*)_.*-.*
- Tier Regex:
(.*)_(.*)_(.*)-.*
Application | Tiers and its VMs |
---|---|
MyApplication1 |
|
MyApplication2 |
|
Discovering Applications Based on Flows
VMware Aria Operations for Networks uses machine learning and statistical analysis to discover applications automatically and group the VMs into their respective applications and tiers. VMware Aria Operations for Networks analyzes the network flow between the VMs and groups the VMs by identifying the application and the tier boundaries between these VMs.
The system analyzes the flows on multiple dimensions such as density of the flows between VMs, common open ports between VMs, number of incoming against outgoing connections, configuration data from load balancers, and so on, and groups VMs with high similarity across these dimensions into a unique application. The process repeats again, but only among the VMs that are already grouped into an application, and to group VMs into their respective tiers.
-
To discover applications based on flows, you must match the following minimum requirements:
- Have at least 10 VMs discovered in VMware Aria Operations for Networks.
- Have at least 100 VM-VM flows.
- Have an Enterprise license.
- Use a platform of XL brick size.
To learn about the configuration requirements, see the System Recommendations and Requirements topic.
If your configuration does not meet the minimum system requirement, you can either scale up your platform cluster or reconfigure your brick size. To learn about how to scale up the platform cluster, see Planning to Scale up the Platform Cluster topic, and to learn about reconfiguring the brick size, see KB-83374.
-
To improve the accuracy of flow-based application discovery, you must add load balancers to VMware Aria Operations for Networks. To learn about the supported load balancers, see Supported Load Balancers.
In VMware Aria Operations for Networks, you can use the Discovery Options feature to upload a CSV file to discover applications. The CSV file can either be exported from the content management database (CMDB) or manually created from other tools to improve the Flow based application and tier discovery. The inputs of the CSV file are verified using the Flow based application discovery algorithm and then used to discover applications. The uploaded CSV file can also be used to name applications and tiers.
This feature can enhance the discovery of applications where only some of the application definitions are known, and the application definitions may not be completely accurate. Flow based application discovery can still add VMs or remove VMs using the inputs of the CSV file depending on the actual Flow pattern at the time of discovery.