Telemetry Data Mapper to map data identifiers from SNMP, gRPC, NETCONF, CLI, etc. to each other.
TDM provides an offline, immutable view into advertised data availability from data models, with search affordance to quickly identify data of interest and the capability to map between data to aid in keeping track of what is roughly equivalent to what.
For example, discovering and mapping...
SNMP OID | YANG XPath |
---|---|
bgpLocalAS |
Cisco-IOS-XR-ipv4-bgp-oper:bgp/instances/instance/instance-active/default-vrf/global-process-info/global/local-as |
ifMTU |
Cisco-IOS-XR-pfi-im-cmd-oper:interfaces/interface-xr/interface/mtu |
cdpCacheDeviceId |
Cisco-IOS-XR-cdp-oper:cdp/nodes/node/neighbors/details/detail/device-id |
... | ... |
In its current state, network telemetry can be accessed in many different ways that are not easily reconciled – for instance, finding the same information in SNMP MIBs and NETCONF/YANG modules. Discovering the datapaths is often tedious and somewhat arcane, and there is no way of determining if the information gathered will have the same values, or which is more accurate than another. Further, the operational methods of deploying this monitoring varies across platforms and implementations. This makes networking monitoring a fragmented ecosystem of inconsistent and unverified data. There needs to be manageability, and cross-domain insight into data availability.
TDM seeks to solve this problem by providing an overlay platform to generically access network telemetry and capability purported to be supported by an OS/release or platform, and create relationships between individual datapaths to demonstrate qualities in consistency, validity, and interoperability. This will be both exposed by UI for human usage, and API for automated usage. TDM will not seek to provide domain-specific manageability, but serve as an overlay insight tool. There is significantly more documentation in doc/.
This has only been lightly evaluated, but we recommend:
These specifications are also semi-dependent on the expected load levels.
ulimit
settings.bash
et al. support.setup.sh
to install docker-compose
for your user.start.sh [http|https]
to start the Docker stack.stop.sh [http|https]
to stop the Docker stack.reset.sh [http|https]
to stop the Docker stack and delete the persisted storage volumes.Ensure there are no port conflicts or change the forwarded ports in docker-compose*.yml
.
git clone https://www.github.com/cisco-ie/tdm.git
# If docker-compose is not installed...
./setup.sh
# Start the stack!
./start.sh [http|https]
# You're good to go :)
# To monitor ETL process...
docker logs -f tdm_etl_1
Once the containers are built and running, it currently takes ~8 hours for all the data to be fully parsed and TDM to be fully available. Progress is visible through the etl
Docker container logs, usually easily accessible via docker logs -f tdm_etl_1
. Once the etl
container has died, TDM is loaded with the current snapshot of data available. Unfortunately there is not a recurring ETL process at this time.
If deploying with HTTPS, an SSL .crt
and .key
must be placed under nginx/
as tdm.cisco.com.crt
and tdm.cisco.com.key
respectively. Usage of a different filename requires changing these filenames in docker-compose.https.yml
and nginx/nginx.https.conf
. This allows components to run over plain HTTP and have NGINX proxy the encryption of public web traffic - which is wonderfully useful for now albeit not as inherently secure.
All ports will be exposed on your Docker host interface. Typically you don't need to worry what this is, assume everything will be available from 127.0.0.1
/localhost
.
80
(HTTP) or 443
(HTTPS) exposes the TDM Web UI.
/goaccess_web.html
exposes website access statistics./goaccess_db.html
exposes ArangoDB access statistics./goaccess_kibana.html
exposes Kibana access statistics.8529
exposes the ArangoDB Web UI and API.5601
exposes Kibana for exploring the TDM Search cache (Elasticsearch).TDM is dual-licensed with Apache License, Version 2.0 applying to its softwares and Community Data License Agreement – Sharing – Version 1.0 applying to data, such as the mappings.
Drew Pletcher, Einar Nilsen-Nygaard, Joe Clarke, Benoit Claise, Charlie Sestito, Glenn Matthews, Michael Ott
UTDM is dead, long live UTDM.