godd is a CLI tool that makes the OvS' output for the datapath flows human readable
It accepts flows information in the standard input and parses it out to a YAML structured output.
godd's primary goal is to make Nuage openvswitch datapath troubleshooting, verification and monitoring easier to human beings.
Despite that I created godd specifically for Nuage flavored OvS, it should work well on a vanilla OvS instances.
You can download and run godd on the OvS device itself:
ovs-dpctl dump-flows -m | ~/godd
Its not always convenient to download the tool to each and every OvS switch, sometimes you already have the flows output that you can transform with godd.
Since the tool expects the data on the standard input you can either cat a file to it:
cat myFlowsData.txt | ~/godd
or simply echo the flows you have:
echo 'recirc_id(0),tunnel(tun_id=0xc882d,src=10.10.99.99,dst=10.10.1.11,ttl=63,flags(-df-csum+key)),in_port(10),skb_mark(0x20000000),eth(src=5a:eb:de:fe:6b:44,dst=00:00:0a:0a:01:0b),eth_type(0x0800),ipv4(src=172.254.0.59,dst=172.254.0.12,proto=17,ttl=63,frag=no),udp(src=19995,dst=19995), packets:4, bytes:360, used:2.418s, actions:set(eth(src=68:54:ed:00:ec:9e,dst=68:54:ed:00:f1:0e)),set(ipv4(dst=172.254.0.12,ttl=62)),14' | ~/godd
The tool is distributed as a single binary for Linux/Mac/Windows OSes.
Check the links for download the latest (v.0.0.1) version for each supported OS. Feel free to correct the download path to someplace in your $PATH if you wish so.
Linux:
curl -kL https://github.com/hellt/godd/releases/download/v0.0.1/godd_linux > ~/godd && chmod a+x ~/godd
Mac:
curl -kL https://github.com/hellt/godd/releases/download/v0.0.1/godd_darwin > ~/godd && chmod a+x ~/godd
Windows:
curl -kL https://github.com/hellt/godd/releases/download/v0.0.1/godd_windows > ~/godd && chmod a+x ~/godd
Despite being nice and output the flow info in a nicely looking YAML (opinionated) godd sports a few additional enhancements:
- easy flow distinction: each flow will be separated with two newlines and a header in green color with the flow seq. number. Since the YAML output is verbose, the ease of flow distinction makes it a breeze to scroll over the flows and find the right one
- grouping related data: the flow data that nests in a common place will be grouped together. I.e. all the
set()actions will be grouped under a singleset:parent, making the result easier to parse with a pair of eyes. - tunnel_id conversion: in Nuage, hex encoded tunnnel ID has a decimal counterpart that is present on the controllers.
goddaugments hex tun_id field with the decimal representation.
There are a lot of improvements that could be made. Some of them are Nuage related, some are of a general kind:
- output in a JSON/table/condensed formats
- receive openflow ports data to unwrap the port names
- add a hash for the 5-tuple data to make a correlation between the flows on a different switches
- and more
The easiest way to contribute is to file an issue with the dump-flow data that godd failed to parse, this will give me more raw data to analyze and make godd more generic.
Roman Dodin - @ntdvps