Customers will request assistance in finding where a DX OpenExplore (wavefront) proxy resides. Typically this is due to lack of documentation or turnover at the customers site. While we do not manage their infrastructures and ultimately it is the customers responsibility to track their machines, this article will help you, help the customer.
While each environment is unique, there are some common areas to look at for potential Source for Proxy Details.
Proxy Chaining, Relaying or Load Balancer,
Browse > Proxy will give you; Hostname, Last Check-in, Status, Version,
Note: When a proxy restarts it will receive a new unique identifier for its proxy ID. The original Proxy ID is not immediately updated so users will see both proxy id's list "Active" for a short time. After the next collection cycle the first ID will change to status "Orphaned".
API Documentation pages - https://<cluster_name>.<domainName>/api-docs/ui/
GET /api/v2/proxy/{id} Get a specific proxy
Using the proxy ID found above you can find the userID of the person/token that deployed the proxy.
"name": "Proxy on proxy.sample",
"id": "######-####-####-#######",
"hostname": "<hostName>",
"userId": "<userID@domainName>",
Note: hard limit of 1000 objects returned is enforced here. Also it will take the API some time to return details on 1000 proxies. If an environment is large and has more than a thousand proxies it's recommended to use offsets to pull smaller subsets lists of the proxies.
Example:
offset = 0 limit = 500 to pull proxy 1 through 500
offset = 500 limit = 500 to pull proxy 501 through 1001
Naming conventions
wavefront-proxy-########-#### - Names like this with a string of number/letters indicate containerized proxies.
These are deployed and managed, through systems such as: Tanzu Kubernetes Grid Integrated Edition (TKGI), Tanzu Mission Control(TMC), Tanzu for Kubernetes Operations (TKO) , Tanzu Application Service (TAS) or opensource k8s. Direct the customer to their Tanzu or opensource k8s management tools to find these proxies.
For example in the Tanzu Ops Manager offers the means to pull metrics from various sources such as TAS, RabbitMQ, and others, sending the data through a proxy installed through the Wavefront Nozzle. Since the proxy is installed and configured in Ops Manager as seen below, this is where your customer would go to find the proxy information.
To find the IP address select the Wavefront Nozzle tile in Ops Manager and then select the STATUS tab to see all of the VMs deployed - Example below shows the proxy and the IP address.
Proxy Chaining, Relaying or Load Balancer,
Proxies with no external access can relay their metrics to other proxies over a TLS connection this is called relaying. See the links below referencing: How to enable TLS connection article below.
In certain circumstances the output from one proxy may need to be directed at another proxy this is called Proxy Chaining. See the links below referencing : How to Chain Proxies.
Many customers use Load Balancers, this is an important tool for network traffic management however it makes it harder to identify the source proxy. Since DXOE has no access to the source proxy address, customers will need to work with their network team to capture and identify network traffic to uncover the source proxy IPs.
Review Spy metrics coming from proxy
Additional information on metrics being ingested may help the customer identify the environment reporting in through the proxy. They can then use tools such as vCenter to search for proxy VMs in their clusters. See the links below referencing Spy process.
Customers Infrastructure team.
Recommend that the customer's internal team use network tools to trace PPS through the network.
Use Wavefront Top or Spy to Investigate Traffic
VMware Aria Operations for Applications Spy to Investigate Metrics per Proxy
Manual Proxy Install on Linux See Sections: Section How to Chain Proxies, and How to enable TLS connection.