0a7df2220c
introduce 'weave stop' which, as well as stopping&removing the weave container also clears the UDP connection tracking for the weave port, without which UDP connections to a quickly (<3min) re-started weave would fail.
Matthias Radestock2014-09-01 16:08:47 +01:00
7dd0bcd1e0
deal with pmtu verification confirmation in forwarder goroutine ...rather than connection goroutine. This is a precurser to smarter pmtu verification, the logic for which is in the forwarder. issue #15Matthias Radestock2014-08-30 23:18:44 +01:00
aef5111af2
switch tcp nc service for bulk of examples since udp nc is full of weirdness
Matthias Radestock2014-08-30 16:30:35 +01:00
731454a406
introduce 'weave expose' which simplifies access to containers from the host, as well as external access.
Matthias Radestock2014-08-30 15:40:50 +01:00
b9be18c386
set up masquerading as part of 'weave launch' thus simplifying external access
Matthias Radestock2014-08-30 14:25:00 +01:00
98257d0df6
switch to ubuntu image for examples It comes with more stuff pre-installed, such as 'nc', so makes experimentation just a little more convenient.
Matthias Radestock2014-08-30 09:59:13 +01:00
3942f19897
Remove boilerplate and condense the http handler down to one function. Also fixed indenting on one line in shell script.
Bryan Boreham2014-08-29 15:26:04 +01:00
753af453ef
fix off-by-14 error in fragmentation This was harmless since it simply meant our fragments were not as long as they could be.
Matthias Radestock2014-08-28 14:17:00 +01:00
4710a59ac8
calculate an initial effectivePMTU That way as soon as the forwarder is live we have a sensible value
Matthias Radestock2014-08-27 21:40:16 +01:00
41d3a6dc7c
log the *effective* pmtu when dropping frames since that is more relevant
Matthias Radestock2014-08-27 21:38:18 +01:00
6f30cf2daf
8 is better than 10 It's quite common for the real pmtu to be 8 less than what was advertised.
Matthias Radestock2014-08-26 21:11:24 +01:00
c5d39fd5bf
drop maxpayload in conjunction with clientPMTU Seems odd to do the latter w/o the former. And wrong.
Matthias Radestock2014-08-26 21:09:23 +01:00
c6a796675d
pay attention to EMSGSIZE errors when sending pmtu verification packets and drop the pmtu accordingly This way, when we actually *do* get pmtu info from the network/kernel we can discover the correct pmtu much more quickly. Especially in the absence of other traffic, forwarding of which was previously the only place where we'd check for EMSGSIZE.
Matthias Radestock2014-08-26 21:05:13 +01:00
fa5a5f46b4
usability: check that script is being run as root so we can display a nicer error if not
Matthias Radestock2014-08-26 13:40:09 +01:00
b35f83f132
explain that we need docker (yes, really) and switch from 'curl' to 'wget' for installing weave, since the latter is more widely installed. Also, make root use during that more explicit.
Matthias Radestock2014-08-26 11:24:47 +01:00
565856eaba
run weave in privileged container ...since on selinux (and possibly others) we encounter a "bind: permission denied" error instead, most likely because weave is attempting to bind a raw ip socket, which is a restricted operation on those systems.
Matthias Radestock2014-08-26 10:27:47 +01:00
516c1363e3
eliminate dependency on pipework ...and docker-ns (though we still leave the script around since it's handy for debugging), and arping (as I have yet to encounter a situation where it was required)
Matthias Radestock2014-08-22 18:14:17 +01:00
7a09d97cf5
when forwarding frames, send icmp 3.4 packets back to the source ...instead of injecting them on the local interface. fixes#2Matthias Radestock2014-08-22 15:18:31 +01:00
7c4464c852
pcap handles are not thread safe they contain an errbuf; there might be other issues too. So we must use different handles in the sniffer and udplistener goroutines.
Matthias Radestock2014-08-20 11:42:55 +01:00
Matthias Radestock