| Commit message (Collapse) | Author | Age | Lines |
|---|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
|
| |
E.g.: For authorization.
We bootstrap the file as empty by default, which means 0 authentication.
|
| | |
|
| |
|
|
|
|
| |
Probably has to be tweaked a bit, but it works.
Fixes #146
|
| | |
|
| | |
|
| |
|
|
|
| |
There really should be a way to make it obey cache-control instead of
this rubbish.
|
| |
|
|
|
| |
Fixes #143
Fixes #141
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
| |
- Adds latency graphs everywhere, even where they shouldn't be. For PoC.
- Fixes a frontend leakage of handlers
- Adds a nasty way of getting latency-data into graphite (needs to be
improved)
- Adds working graphite integration in varnish
- Adds config for graphite
Lots of things broken here, but it's a half-decent start.
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
|
|
|
|
| |
When a device replies from two or more IPs we can't assume it actually
knows what IPs it replied from... (great, right?).
This adds a fall-back when that happens where we compare the actual
interfaces. We require a minimum of 10 matching MAC addresses and skip the
once that are blank to gain confidence in the result.
|
| |
|
|
|
|
|
|
|
|
|
|
| |
Truns out people suck at making sensible network hardware.
Lots of replies on IPs that the devices didn't know they had. I discard
these now, but should keep some sort of map. The reason this is a problem
is that we later "Deduplicate" results by comparing IP lists - but that
doesn't work when the devices don't know about all of the IPs they listen
to.
I might have to revisit this since the IPs are obviously valid...
|
| |
|
|
|
| |
Now also populates snmpresults artificially with data from LLDP when
needed.
|
| |
|
|
|
|
|
| |
Still need to populate the ipmap with results from LLDP somehow.
E.g.: If we can't contact a switch over SNMP that we see over LLDP we know
the sysname and IP.
|
| |
|
|
|
|
|
| |
This removes the actual mapping, but introdcues crawling.
All the data for mapping is there, and now the collision detection and
fixing is starting to look actually sensible.
|
| | |
|
| | |
|
| |
|
|
|
| |
The lldp-only map ensures we aren't fooled by vlans and other logical
links, but if lldp is disabled it obviously doesn't tell the whole story.
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
what?
lol...
just figure it out for me.
Ok, so the tool isn't entirely done, but it departs from the simple
lldpdiscver.pl method that was very limited when things weren't working as
they should.
The long-term idea of lolwhat is that it should be able to do all sorts of
discovery. Right now it just takes a list of IPs and figures out how things
are connected by SNMP. It uses LLDP info from SNMP, but it does not
_currently_ traverse new hosts based on what it finds.
|
| |
|
|
|
|
|
|
| |
Still needs lots of work, but mainly with regards to gluing it all
together.
Possibly populate a nice little "local macs" hash to map mac addresses to
systems/ports easily.
|
| | |
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
As a result of testing with TG16-data, I was able to find and fix a few
outstanding issues and tune the health map to make more sense.
Also introduced both temperature and dhcp into the health map and info box.
Removed the switchInfo text on the health map unless the score is above a
threshold.
Still more tuning needed. After introducing dhcpInfo it's become even more
important to address #88
|
| |
|
|
|
|
| |
References #53 #52 #54
It's hardcoded, but I can live with that for now.
|
| |\ |
|
| | | |
|
| |/ |
|
| | |
|
| | |
|
| | |
|
| |
|
|
| |
Still lots to do
|
| | |
|
| |
|
|
| |
Somewhat dirty pwd-hack
|
| | |
|
| |
|
|
|
|
|
|
| |
Never been so happy to see a single switch in the nms.
Missing: Varnish and collectors and better deplist.
Also: this is mainly _test_ stuff, not meant for development OR production.
|
| | |
|
| | |
|
|
|
moving things that need work into extras/
|
