Target Access Point Ambassador (TAPA)
Description
The Target Access Point Ambassador (TAPA) is the main interface user applications can use to interact with Meridio. The TAPA is a local service running in the user pod as a side-car container. User applications can connect to the TAPA server via a unix socket file shared via a volume. Once connected to the server, user applications can consume the TAPA API via three different function: Open/Close/Watch. The proto file of the TAPA can be found here.
The watch function allows user applications to monitor the streams opened on the pod and check their status.
Open/Close handling
The open process takes place in 3 steps:
- Connect to the trench if not already connected.
- Connect to the conduit if not already connected.
- Register to target registry if not already registered.
While a conduit is opened, the TAPA will monitor the VIP addresses in the trench via the NSP and update the connection to NSM by including the VIPs addresses in the IP Context. Thus, NSM will update the connection and add/remove IP addresses from the conduit target network interface.
The close process takes place in 3 steps:
- Unregister itself from the NSP Target registry.
- Close the conduit if no more stream opened in the conduit.
- Disconnect from the trench (disconnect from NSP) if no more conduit opened. If for some reason
The only error the TAPA can return is to indicate to the client that the pod is already connected to another trench. Indeed, a pod cannot be connected to more than one trench at a time.
Registration
TODO
Refresh
While a stream is opened, the TAPA has to refresh its registration to the target registry to not be considered as a "ghost" entry and then to not be removed automatically.
To achieve the refresh, the TAPA will update its target with the same properties by calling the Register function regularly. After the call, the TAPA will check if the target is still in the ENABLED
status for the target registry. If not, it means the target registry has previously removed the target and consider the last register call as a new entry, so it has overwritten the status to DISABLED
. The TAPA will then start the registeration process a new time to register itself with a unique identifier.
Configuration
https://github.com/Nordix/Meridio/blob/master/cmd/tapa/config.go
Environment variable | Type | Description | Default |
---|---|---|---|
MERIDIO_NAME | string | Name of the pod | nsc |
MERIDIO_NODE | string | Node name the target is running on | |
MERIDIO_NAMESPACE | string | Namespace the trenches to connect to are running on | default |
MERIDIO_SOCKET | string | Path of the socket file of the TAPA | /ambassador.sock |
MERIDIO_NSM_SOCKET | url.URL | Path of the socket file of NSM | unix:///var/lib/networkservicemesh/nsm.io.sock |
MERIDIO_NSP_SERVICE_NAME | string | Domain name of the NSP Service | nsp-service |
MERIDIO_NSP_SERVICE_PORT | int | port of the NSP Service | 7778 |
MERIDIO_TIMEOUT | time.Duration | timeout of NSM request/close, NSP register/unregister... | 15s |
MERIDIO_DIAL_TIMEOUT | time.Duration | timeout to dial NSMgr | 5s |
MERIDIO_MAX_TOKEN_LIFETIME | time.Duration | maximum lifetime of tokens | 24h |
MERIDIO_LOG_LEVEL | string | Log level | DEBUG |
MERIDIO_NSP_ETRY_TIMEOUT | time.Duration | Timeout of the entries registered in NSP | 30s |
MERIDO_GRPC_MAX_BACKOFF | time.Duration | Upper bound on gRPC connection backoff delay | 5s |
MERIDIO_GRPC_PROBE_RPC_TIMEOUT | time.Duration | RPC timeout of internal gRPC health probes if any | 1s |
Command Line
Command | Action | Default |
---|---|---|
--help | Display a help describing | |
--version | Display the version | |
--debug | Prints meridio-version, unix-time, network-interfaces, rules, route, neighbors, system information, and environment-variables in a json format |
Communication
Here are all components the tapa is communicating with:
Component | Secured | Method | Description |
---|---|---|---|
Spire | TBD | Unix Socket | Obtain and validate SVIDs |
NSM | yes (mTLS) | Unix Socket | Request/Close/Watch connections |
NSP Service | yes (mTLS) | TCP | Watch configuration. Register/Unregister targets |
User Application | no | Unix Socket | Provide TAPA service |
An overview of the communications between all components is available here.
Health check
The health check is provided by the GRPC Health Checking Protocol. The status returned can be UNKNOWN
, SERVING
, NOT_SERVING
or SERVICE_UNKNOWN
.
Service | Description |
---|---|
Liveness | A unique service to be used by liveness probe to return status, can aggregate other lesser services |
Startup | A unique service to be used by startup probe to return status, can aggregate other lesser services |
Readiness | A unique service to be used by readiness probe to return status, can aggregate other lesser services |
Service | Probe | Description |
---|---|---|
AmbassadorSvc | Liveness | Monitor status of the Ambassador server |
Privileges
No privileges required.