8cf30395a1
Changed default bridge interface do 'docker0'
2013-04-05 14:16:19 -07:00
793c1ad990
Merge remote-tracking branch 'origin/219-default-bridge-2'
2013-04-05 14:02:16 -07:00
febaeebfb8
Add tests of tcp port allocator
2013-04-05 13:03:24 -07:00
d32f184696
Fix a race condition when running the port allocator
2013-04-05 13:03:04 -07:00
2aad4a3478
Choose which TCP frontend port to allocate with '-p :PORT'
2013-04-04 22:58:01 -07:00
a5fb1d6c01
Refactored PortAllocator to allow for same-frontend constraint
2013-04-04 22:56:12 -07:00
f344212b93
Renamed PortAllocator.populate() to run()
2013-04-04 19:49:32 -07:00
0424998f38
Print a less confusing error message when lxcbr0 doesn't exist
2013-04-04 19:14:10 -07:00
586a79cca0
Merge remote-tracking branch 'dominikh/minor-code-touchups'
2013-04-04 16:20:37 -07:00
ebc837957f
Continue cleaning up iptables rules from previous version, to avoid crashing after an upgrade
2013-04-04 15:16:42 -07:00
32f5811476
stop looping remote:port from host to containers
2013-04-04 23:07:10 +03:00
1b370f9d8d
Move the default bridge name to a constant
2013-04-04 05:33:28 -07:00
aa4bf4284b
If bridge does not exists, try to create it
2013-04-03 16:17:03 -07:00
90a6e310fe
Add an helper function to check if two network overlaps. Also add unit tests for this function
2013-04-03 16:15:43 -07:00
f39af7e05d
Put the bridge interface name in the command line
2013-04-03 16:15:43 -07:00
3b65be9127
Fix NAT problem with ports looping back to containers
2013-04-04 01:32:46 +03:00
22f1cc955d
replace unreachable returns with panics
...
Not only is this a more common idiom, it'll make finding bugs easier,
and it'll make porting to Go 1.1 easier.
Go 1.1 will not require the final return or panic because it has a
notion of terminating statements.
2013-04-03 11:18:23 +02:00
6f9a67a7c7
Make IP allocator lazy
...
Instead of allocating all possible IPs in advance, generate them as
needed.
A loop will cycle through all possible IPs in sequential order,
allocating them as needed and marking them as in use. Once the loop
exhausts all IPs, it will wrap back to the beginning. IPs that are
already in use will be skipped. When an IP is released, it will be
cleared and be available for allocation again.
Two decisions went into this design:
1) Minimize memory footprint by only allocating IPs that are actually
in use
2) Minimize reuse of released IP addresses to avoid sending traffic to
the wrong containers
As a side effect, the functions for IP/Mask<->int conversion have been
rewritten to never be able to fail in order to reduce the amount of
error returns.
Fixes gh-231
2013-04-01 06:02:44 +02:00
523803d633
Handling iptables() errors more usefully during portmapper setup
2013-03-28 14:44:54 -05:00
c66d2b6a53
Return error when iptables is not found
2013-03-28 14:30:56 -05:00
dfc3904f77
Looking for iptables in PATH
2013-03-28 14:02:50 -05:00
51455b1ee0
Merge pull request #138 from srid/handle-port-mapper-error
...
handle errors during the creation of port mapper
2013-03-24 10:27:03 -07:00
371225520f
handle errors during the creation of port mapper
...
example:
2013/03/22 21:42:55 Unable to setup port networking: Failed to create DOCKER chain
(which was possibly introduced by commit 3c6b8bb88
2013-03-22 21:44:01 -07:00
301a8afff5
Properly cleanup iptables rules inserted in OUTPUT (introduced in 3c6b8bb888)
2013-03-22 22:31:20 -07:00
3c6b8bb888
Fixing Issue #98 : Adding DOCKER to output chain during iptables setup
2013-03-22 11:28:15 -05:00
3aefed2dc2
When lxcbr0 has several associated IPs, default to first one found
2013-03-21 09:19:22 -07:00
ab99e9252d
Complete pull request #121 , init TCPAddr with named field
2013-03-20 06:02:25 -07:00
fac32cda5a
Fix issue #120 , initialize TCPAddr w/ field names
...
Current Go tip (+74e65f07a0c8) and likely Go 1.1 does not build docker since net.TCPAddr struct has an additional field now for IPv6:
type TCPAddr struct {
IP IP
Port int
Zone string // IPv6 scoped addressing zone
}
Initializing the struct with named fields resolves this problem.
2013-03-21 00:11:16 -03:00
799ffa1763
Network: Port mapping support.
...
Implemented a port allocator and a port mapper that is able to forward TCP ports
from the host to the container.
2013-02-28 11:50:02 -08:00
c08f5b2b84
Integrated the network allocator into Docker. A networking environment
...
is assigned to each container upon Start and released whenever the
container exits.
2013-02-25 14:06:22 -08:00
797bb6e75b
Network allocator
2013-02-25 10:45:23 -08:00
e0e49b9a22
Network: Do not assume that we are using a class C. Instead, compute the
...
IP addresses range and network size in order to allocate an IP address.
2013-02-21 18:33:23 -08:00
6124c5eb31
Network: Simple random IP allocation on the bridge network.
2013-02-21 10:47:57 -08:00
5039d4a280
Network: Automatically figure out the gateway and netmask by inspecting
...
the lxc bridge interface
2013-02-21 10:47:57 -08:00
5cecd548cd
Basic networking support with hardcoded addresses. Work in progress.
2013-02-21 10:47:57 -08:00
Solomon Hykes