Commit Graph

91 Commits

Author SHA1 Message Date
Brian Tiger Chow
a85aa9b5da refactor: rename testutil func 2014-12-24 09:31:17 -05:00
Juan Batiz-Benet
e630ca6078 dht: removing extra newlines 2014-12-24 03:42:43 -08:00
Juan Batiz-Benet
ccf6f79aa0 respect don contexteone 2014-12-24 03:24:28 -08:00
Juan Batiz-Benet
c84a714b16 peer change: peer.Peer -> peer.ID
this is a major refactor of the entire codebase
it changes the monolithic peer.Peer into using
a peer.ID and a peer.Peerstore.

Other changes:
- removed handshake3.
-	testutil vastly simplified peer
-	secio bugfix + debugging logs
-	testutil: RandKeyPair
-	backpressure bugfix: w.o.w.
-	peer: added hex enc/dec
-	peer: added a PeerInfo struct
  PeerInfo is a small struct used to pass around a peer with
 	a set of addresses and keys. This is not meant to be a
 	complete view of the system, but rather to model updates to
 	the peerstore. It is used by things like the routing system.
-	updated peer/queue + peerset
-	latency metrics
-	testutil: use crand for PeerID gen
 	RandPeerID generates random "valid" peer IDs. it does not
 	NEED to generate keys because it is as if we lost the key
 	right away. fine to read some randomness and hash it. to
 	generate proper keys and an ID, use:
 	  sk, pk, _ := testutil.RandKeyPair()
 	  id, _ := peer.IDFromPublicKey(pk)
 	Also added RandPeerIDFatal helper
- removed old spipe
- updated seccat
- core: cleanup initIdentity
- removed old getFromPeerList
2014-12-23 08:33:32 -08:00
Juan Batiz-Benet
ba323c3e48 make vendor 2014-12-17 23:25:38 -08:00
Juan Batiz-Benet
6334e19374 make net work with new stream + mux 2014-12-16 05:10:46 -08:00
Juan Batiz-Benet
6bc26f1752 writing swarm2 using go-peerstream
omg wow such pass
2014-12-16 05:10:46 -08:00
Juan Batiz-Benet
a3a48ce67b conn: tests pass :) 2014-12-16 05:10:45 -08:00
Juan Batiz-Benet
393842e245 much simpler net
- removed ctxcloser
- removed multiconn
- focused on netio
2014-12-16 05:10:45 -08:00
Juan Batiz-Benet
9b7fb704ef make net/conn.Conn conform to net.Conn 2014-12-16 05:10:45 -08:00
Juan Batiz-Benet
3b29653022 updated msgio + secio 2014-12-16 05:10:45 -08:00
Juan Batiz-Benet
129eca0d82 multiconn: close fanIn + error 2014-12-16 05:10:44 -08:00
Juan Batiz-Benet
5341379f9d net/conn: io, not channels
This commit changes the connections to use io.ReadWriters
instead of channels (+ async workers). This is a pretty
big change -- away from csp -- in the name of performance
(and predictable flow control).

It also uses the brand new secio, which is spipe's successor.
2014-12-16 05:10:44 -08:00
Brian Tiger Chow
93fb9fb929 style(logging) demote some statements to Debug (from info)
to improve SnR. Feel free to contest these changes if you find these
statements useful. In my experience, they were adding noise at Info.

License: MIT
Signed-off-by: Brian Tiger Chow <brian@perfmode.com>
2014-12-08 15:07:33 -08:00
Brian Tiger Chow
c67286774c silence verbose output
for higher SnR at IPFS_LOGGING=info

License: MIT
Signed-off-by: Brian Tiger Chow <brian@perfmode.com>
2014-12-08 15:07:33 -08:00
Brian Tiger Chow
74eb03231d fix(core, peer) helpers to testutil, err handling
License: MIT
Signed-off-by: Brian Tiger Chow <brian@perfmode.com>
2014-12-08 14:32:52 -08:00
Brian Tiger Chow
d77c4bb5e0 refactor(peer): create peer through peerstore
for safety!

use mockpeer.WithID methods to create peers in tests

License: MIT
Signed-off-by: Brian Tiger Chow <brian@perfmode.com>
2014-12-08 02:10:42 -08:00
Juan Batiz-Benet
59c1ab2fc3 Merge pull request #406 from jbenet/new-msgio
updated msgio
2014-12-06 16:20:51 -08:00
Juan Batiz-Benet
a522a5a532 conn: larger max message.
With the new msgio, we no longer need to bound message sizes for
the sake of buffers. MaxMessageSize should remain though in order
to prevent streaming of huge things. In general, messages shouldn't
be above one MB. This max is now set to 4MB.

If a block happens to be larger, we should split it up.
2014-12-06 11:29:31 -05:00
Juan Batiz-Benet
97e0f73e79 msgio: updated to remove max size 2014-12-06 11:26:47 -05:00
Brian Tiger Chow
f870948274 fix: multiconn s/Conns()/getConns()
@jbenet

must be getConns to avoid clash with private var

License: MIT
Signed-off-by: Brian Tiger Chow <brian@perfmode.com>
2014-12-05 20:56:28 -08:00
Brian Tiger Chow
229c93f72f fix(net/multiconn) data race in test
https://build.protocol-dev.com/job/go-ipfs.test.go.race.nofuse/276/console

License: MIT
Signed-off-by: Brian Tiger Chow <brian@perfmode.com>
2014-12-05 20:56:26 -08:00
Jeromy
34da5f039c log formatting (info -> infof) 2014-11-21 03:54:26 +00:00
Brian Tiger Chow
93cf2bb176 style(net/conn) fmt -> debugerror
License: MIT
Signed-off-by: Brian Tiger Chow <brian@perfmode.com>
2014-11-20 16:14:23 -08:00
Juan Batiz-Benet
48879dd56c conn: added (*dialer).DialAddr 2014-11-20 09:14:49 -08:00
Juan Batiz-Benet
e0dab75d54 swarm: better err on listen fail 2014-11-20 09:14:48 -08:00
Juan Batiz-Benet
fc6bab7b73 net/conn: moved zero addr error 2014-11-20 01:27:18 -08:00
Juan Batiz-Benet
43452fdeb7 updated multiaddr + multiaddr-net
I broke out multiaddr/net to its own package so
that multiaddr parsing could remain lean. multiaddr-net
will vendor special implementations of protocols (like utp)
2014-11-20 00:58:48 -08:00
Brian Tiger Chow
6291c25455 fix(net/conn) Set peer to "Remote" in Dial method
@jbenet we may want to consider setting the default type to Remote. I
left the default as Unspecified just for safety.

License: MIT
Signed-off-by: Brian Tiger Chow <brian@perfmode.com>
2014-11-16 07:01:01 -08:00
Brian Tiger Chow
ed4a8eb782 chore(tests) add Short() -> SkipNow() to slowest tests
vanilla:
    21.57 real        45.14 user         8.51 sys

short:
    14.40 real        31.13 user         5.56 sys

License: MIT
Signed-off-by: Brian Tiger Chow <brian@perfmode.com>
2014-11-15 02:10:53 -08:00
Jeromy
b0f4b7c89d some net docs 2014-11-08 21:42:36 -08:00
Juan Batiz-Benet
3bc2281c51 fixed race in multiconn 2014-11-05 09:44:51 -08:00
Juan Batiz-Benet
d26fd58182 ctx closer races #270 2014-11-05 09:26:17 -08:00
Juan Batiz-Benet
4989dcafed swarm + handshake: better observed addr check
The check needed knowledge of the _listen_ addresses,
not just the interface addresses. Also, the handshake now
sends out all the addresses we accumulate about ourselves.
(this may be bad in the long run, but useful now to test)
2014-11-05 04:01:38 -08:00
Brian Tiger Chow
98ebe2f3ec hotfix(vendor) 2014-11-04 08:48:27 -08:00
Juan Batiz-Benet
d303ff45f8 Merge pull request #253 from jbenet/net-detect
NAT detect
2014-11-03 16:53:09 -08:00
Jeromy
3e62042745 fix up NAT check and clear up naming 2014-11-03 15:10:44 -08:00
Jeromy
88bf39118c comment comment comment comment 2014-11-03 03:02:56 +00:00
Jeromy
402fdd2db9 changes from PR 2014-11-03 01:45:45 +00:00
Jeromy
094baf6fbf error out if attempting connection to loopback 2014-11-03 00:28:07 +00:00
Jeromy
cb2fb9cfe6 print NAT if detected 2014-11-02 21:59:59 +00:00
Jeromy
2eac5aeec6 begin code for NAT detection 2014-11-02 21:19:22 +00:00
Jeromy
a8069024e9 cut down on allocations 2014-11-01 16:07:56 -07:00
Jeromy
027920540f more memory tweaks 2014-11-01 16:07:56 -07:00
Jeromy
9b1c99e9aa shrink msgio buffer size to decrease memory pressure 2014-11-01 16:07:56 -07:00
Jeromy
f89cfc1108 moved pipes package 2014-11-01 16:07:56 -07:00
Jeromy
07733b17b3 msgio pooling first hack 2014-11-01 16:07:56 -07:00
Jeromy
ffc59fff1b add blowfish code and refactor pipes and duplex a little 2014-11-01 16:07:56 -07:00
Brian Tiger Chow
cea398b8a7 fix(all) log.Debug -> log.Debugf 2014-10-30 06:35:29 -07:00
Jeromy
d92db12460 lots of logging 2014-10-26 00:45:40 +00:00