Matthias Radestock
83464a38eb
remove explicit multicast detection
...
We don't need. Multicast addresses never appear as sources and hence
never end up in the peer_cache. And the router always broadcasts
frames with destination addresses that aren't in the peer cache.
2014-10-24 14:30:58 +01:00
David Wragg
ded7d3a991
Merge pull request #152 from rade/weave_script_tidy
...
various simplifications and tidying up in weave script
2014-10-24 14:26:35 +01:00
Matthias Radestock
dba8760187
update README
2014-10-24 13:43:15 +01:00
Bryan Boreham
5ee60a0003
Extend tests slightly
2014-10-24 13:37:32 +01:00
Matthias Radestock
00a4a5e056
refactor: simplify multicast route addition in 'attach'
2014-10-24 13:15:15 +01:00
Matthias Radestock
560cbabb54
refactor: simplify 'attach'
...
Move the interface into the container network NS asap, which
simplifies error handling.
2014-10-24 13:13:34 +01:00
Matthias Radestock
de96742935
refactor: remove add_nat_rule
...
it did too little to be worthwhile
2014-10-24 13:03:28 +01:00
Bryan Boreham
fa0331cb6d
Ensure we only attempt to handle TypeA queries
2014-10-24 13:03:26 +01:00
Matthias Radestock
093b23fa83
refactor: it's not an IPADDR but a CIDR
2014-10-24 12:59:47 +01:00
Matthias Radestock
5b6a892c38
refactor: don't extract param to var and shift unless we have to
2014-10-24 12:57:35 +01:00
Matthias Radestock
6501934a15
make with_container_netns take the container as the first arg
...
and invoke the subsequent command.
This is closer to typical 'with' semantics.
2014-10-24 12:50:29 +01:00
Matthias Radestock
34e240c4a5
remove misleading comment...
...
attach() and detach() are never invoked directly,
plus it's really not that hard to figure out what they do.
2014-10-24 12:49:04 +01:00
Bryan Boreham
d6f313eac2
Improve debug messages
2014-10-24 12:46:02 +01:00
Bryan Boreham
5f3080b9db
Don't crash when printing name of nil interface
2014-10-24 12:37:28 +01:00
Matthias Radestock
579f3469c8
simplify code for attach
2014-10-24 12:30:42 +01:00
Bryan Boreham
9b90e119e5
Say 'name' instead of 'identity' in log and status
2014-10-24 11:38:10 +01:00
Matthias Radestock
7a53efe66c
get error handling right
2014-10-24 11:27:32 +01:00
Matthias Radestock
73f504afd6
Merge branch '130_launch_without_ip' of github.com:rade/weave into 130_launch_without_ip
...
Conflicts:
weave
2014-10-24 10:23:50 +01:00
Matthias Radestock
85787f2d4e
fix 'weave start' post rebase
2014-10-24 10:21:02 +01:00
Matthias Radestock
0adda9eae9
drop IP/network from 'weave launch'
...
plus a whole bunch of other small changes to weave script which I'll
eventually extract into separate commits.
2014-10-24 10:19:22 +01:00
Bryan Boreham
ced7299140
Removed weave ps -v option
2014-10-24 10:14:32 +01:00
Bryan Boreham
e4fe57a02c
Make extra container information optional
2014-10-24 10:14:32 +01:00
Bryan Boreham
f7eeb1f844
Fix issues raised in code review
2014-10-24 10:14:32 +01:00
Bryan Boreham
06adae8fb5
Added weave ps command
2014-10-24 10:14:32 +01:00
Matthias Radestock
46b0cc35f1
drop IP/network from 'weave launch'
...
plus a whole bunch of other small changes to weave script which I'll
eventually extract into separate commits.
2014-10-24 10:11:43 +01:00
Bryan Boreham
988f555a68
Merge pull request #150 from rade/126_weave_start
...
LGTM. Closes #126 , also referenced from #148
2014-10-24 09:30:07 +01:00
Matthias Radestock
070ec0c0e9
add usage
2014-10-23 18:00:29 +01:00
Matthias Radestock
9cc6f9121e
add 'weave start'
2014-10-23 17:55:23 +01:00
Matthias Radestock
041a4c5f6c
don't talk about netmasks
2014-10-23 17:50:38 +01:00
Matthias Radestock
f037865b82
consistency
2014-10-23 17:27:35 +01:00
Matthias Radestock
7778946f96
be more correct / precise about CIDR notation in usage
2014-10-23 17:25:34 +01:00
Michael Bridgen
ad44a8816f
Remove subnets from API, since we don't use them
2014-10-23 16:43:41 +01:00
Michael Bridgen
840c0b9f86
Renamed variables for consistency
2014-10-23 16:23:55 +01:00
Matthias Radestock
007f78e07c
make 'docker load -i weave.tar' work again
...
for some reason it started complaining about an unknown ids for me
2014-10-23 12:09:41 +01:00
Bryan Boreham
c63b163703
Formatting in error message
2014-10-22 18:27:55 +01:00
Bryan Boreham
94a812d98f
Factor out error type checks
2014-10-22 18:24:02 +01:00
Bryan Boreham
0a5990bd6e
Factor out testing status checks
2014-10-22 16:59:28 +01:00
Bryan Boreham
d84226cc31
Factor out testing error checks
2014-10-22 16:52:35 +01:00
Matthias Radestock
eb5548b1c3
move docker-ns
2014-10-22 14:29:18 +01:00
Matthias Radestock
b8c308dd09
move weave script
2014-10-22 14:14:33 +01:00
Matthias Radestock
8dd97a07de
fix up Makefile to cope with its new location
...
also various other Makefile tidy-ups
- extract consts for various targets
- mark non-file targets as phony
- use $@ when possible
2014-10-22 14:06:16 +01:00
Matthias Radestock
bc02066bdb
move Makefile to top-level
2014-10-22 13:36:06 +01:00
Michael Bridgen
156314a802
Complete multicast requests on first response
...
This is the simplest thing that works, is consistent with what happens
if there's a local answer, and avoids having all multicast requests
last 250ms or more.
Since we're no longer waiting *at least* the timeout period, I've
bumped the multicast response timeout up to half a second.
There was also a bug whereby the query would be entered into the
inflight map even if the socket write had failed (and the channel had
been closed).
2014-10-22 13:35:40 +01:00
Michael Bridgen
e21069f3ab
Factored out HTTP error + log
2014-10-22 11:58:55 +01:00
Michael Bridgen
1d5a55ea75
Factored out more (but different) err != nil code
2014-10-21 17:06:46 +01:00
Michael Bridgen
0615d122e9
Made zone return errors and http idempotent
...
The Zone DB tells you if you have a duplicate or missing record, but
the HTTP server smoothes these over.
2014-10-21 17:05:03 +01:00
Michael Bridgen
5c6410a3c6
Refactored indexOf*
2014-10-21 15:49:42 +01:00
Michael Bridgen
23edd79177
Changed Lock to RLock when matching
2014-10-21 14:01:07 +01:00
Michael Bridgen
6eb7d71499
Substituted checkFatal for repeated if err == nil
2014-10-21 13:58:12 +01:00
Michael Bridgen
1a593fc81f
Used an init func instead of initialisers
...
Ends up slightly neater. (Also note Error goes from os.Stdout to
os.Stderr)
2014-10-21 12:25:05 +01:00
Matthias Radestock