Mercurial > hg
comparison tests/test-http-protocol.t @ 30764:e75463e3179f
protocol: send application/mercurial-0.2 responses to capable clients
With this commit, the HTTP transport now parses the X-HgProto-<N>
header to determine what media type and compression engine to use for
responses. So far, we only compress responses that are already being
compressed with zlib today (stream response types to specific
commands). We can expand things to cover additional response types
later.
The practical side-effect of this commit is that non-zlib compression
engines will be used if both ends support them. This means if both
ends have zstd support, zstd - not zlib - will be used to compress
data!
When cloning the mozilla-unified repository between a local HTTP
server and client, the benefits of non-zlib compression are quite
noticeable:
engine server CPU (s) client CPU (s) bundle size
zlib (l=6) 174.1 283.2 1,148,547,026
zstd (l=1) 99.2 267.3 1,127,513,841
zstd (l=3) 103.1 266.9 1,018,861,363
zstd (l=7) 128.3 269.7 919,190,278
zstd (l=10) 162.0 - 894,547,179
none 95.3 277.2 4,097,566,064
The default zstd compression level is 3. So if you deploy zstd
capable Mercurial to your clients and servers and CPU time on
your server is dominated by "getbundle" requests (clients cloning
and pulling) - and my experience at Mozilla tells me this is often
the case - this commit could drastically reduce your server-side
CPU usage *and* save on bandwidth costs!
Another benefit of this change is that server operators can install
*any* compression engine. While it isn't enabled by default, the
"none" compression engine can now be used to disable wire protocol
compression completely. Previously, commands like "getbundle" always
zlib compressed output, adding considerable overhead to generating
responses. If you are on a high speed network and your server is under
high load, it might be advantageous to trade bandwidth for CPU.
Although, zstd at level 1 doesn't use that much CPU, so I'm not
convinced that disabling compression wholesale is worthwhile. And, my
data seems to indicate a slow down on the client without compression.
I suspect this is due to a lack of buffering resulting in an increase
in socket read() calls and/or the fact we're transferring an extra 3 GB
of data (parsing HTTP chunked transfer and processing extra TCP packets
can add up). This is definitely worth investigating and optimizing. But
since the "none" compressor isn't enabled by default, I'm inclined to
punt on this issue.
This commit introduces tons of tests. Some of these should arguably
have been implemented on previous commits. But it was difficult to
test without the server functionality in place.
author | Gregory Szorc <gregory.szorc@gmail.com> |
---|---|
date | Sat, 24 Dec 2016 15:29:32 -0700 |
parents | 35b516f800e0 |
children | a1dd2c0c479e |
comparison
equal
deleted
inserted
replaced
30763:a520aefb96f1 | 30764:e75463e3179f |
---|---|
40 $ cat hg.pid > $DAEMON_PIDS | 40 $ cat hg.pid > $DAEMON_PIDS |
41 $ get-with-headers.py 127.0.0.1:$HGPORT '?cmd=capabilities' | tr ' ' '\n' | grep compression | 41 $ get-with-headers.py 127.0.0.1:$HGPORT '?cmd=capabilities' | tr ' ' '\n' | grep compression |
42 compression=none,zlib | 42 compression=none,zlib |
43 | 43 |
44 $ killdaemons.py | 44 $ killdaemons.py |
45 | |
46 Start a default server again | |
47 | |
48 $ hg -R server serve -p $HGPORT -d --pid-file hg.pid | |
49 $ cat hg.pid > $DAEMON_PIDS | |
50 | |
51 Server should send application/mercurial-0.1 to clients if no Accept is used | |
52 | |
53 $ get-with-headers.py --headeronly 127.0.0.1:$HGPORT '?cmd=getbundle&heads=e93700bd72895c5addab234c56d4024b487a362f&common=0000000000000000000000000000000000000000' - | |
54 200 Script output follows | |
55 content-type: application/mercurial-0.1 | |
56 date: * (glob) | |
57 server: * (glob) | |
58 transfer-encoding: chunked | |
59 | |
60 Server should send application/mercurial-0.1 when client says it wants it | |
61 | |
62 $ get-with-headers.py --hgproto '0.1' --headeronly 127.0.0.1:$HGPORT '?cmd=getbundle&heads=e93700bd72895c5addab234c56d4024b487a362f&common=0000000000000000000000000000000000000000' - | |
63 200 Script output follows | |
64 content-type: application/mercurial-0.1 | |
65 date: * (glob) | |
66 server: * (glob) | |
67 transfer-encoding: chunked | |
68 | |
69 Server should send application/mercurial-0.2 when client says it wants it | |
70 | |
71 $ get-with-headers.py --hgproto '0.2' --headeronly 127.0.0.1:$HGPORT '?cmd=getbundle&heads=e93700bd72895c5addab234c56d4024b487a362f&common=0000000000000000000000000000000000000000' - | |
72 200 Script output follows | |
73 content-type: application/mercurial-0.2 | |
74 date: * (glob) | |
75 server: * (glob) | |
76 transfer-encoding: chunked | |
77 | |
78 $ get-with-headers.py --hgproto '0.1 0.2' --headeronly 127.0.0.1:$HGPORT '?cmd=getbundle&heads=e93700bd72895c5addab234c56d4024b487a362f&common=0000000000000000000000000000000000000000' - | |
79 200 Script output follows | |
80 content-type: application/mercurial-0.2 | |
81 date: * (glob) | |
82 server: * (glob) | |
83 transfer-encoding: chunked | |
84 | |
85 Requesting a compression format that server doesn't support results will fall back to 0.1 | |
86 | |
87 $ get-with-headers.py --hgproto '0.2 comp=aa' --headeronly 127.0.0.1:$HGPORT '?cmd=getbundle&heads=e93700bd72895c5addab234c56d4024b487a362f&common=0000000000000000000000000000000000000000' - | |
88 200 Script output follows | |
89 content-type: application/mercurial-0.1 | |
90 date: * (glob) | |
91 server: * (glob) | |
92 transfer-encoding: chunked | |
93 | |
94 #if zstd | |
95 zstd is used if available | |
96 | |
97 $ get-with-headers.py --hgproto '0.2 comp=zstd' 127.0.0.1:$HGPORT '?cmd=getbundle&heads=e93700bd72895c5addab234c56d4024b487a362f&common=0000000000000000000000000000000000000000' > resp | |
98 $ f --size --hexdump --bytes 36 --sha1 resp | |
99 resp: size=248, sha1=4d8d8f87fb82bd542ce52881fdc94f850748 | |
100 0000: 32 30 30 20 53 63 72 69 70 74 20 6f 75 74 70 75 |200 Script outpu| | |
101 0010: 74 20 66 6f 6c 6c 6f 77 73 0a 0a 04 7a 73 74 64 |t follows...zstd| | |
102 0020: 28 b5 2f fd |(./.| | |
103 | |
104 #endif | |
105 | |
106 application/mercurial-0.2 is not yet used on non-streaming responses | |
107 | |
108 $ get-with-headers.py --hgproto '0.2' 127.0.0.1:$HGPORT '?cmd=heads' - | |
109 200 Script output follows | |
110 content-length: 41 | |
111 content-type: application/mercurial-0.1 | |
112 date: * (glob) | |
113 server: * (glob) | |
114 | |
115 e93700bd72895c5addab234c56d4024b487a362f | |
116 | |
117 Now test protocol preference usage | |
118 | |
119 $ killdaemons.py | |
120 $ hg --config server.compressionengines=none,zlib -R server serve -p $HGPORT -d --pid-file hg.pid | |
121 $ cat hg.pid > $DAEMON_PIDS | |
122 | |
123 No Accept will send 0.1+zlib, even though "none" is preferred b/c "none" isn't supported on 0.1 | |
124 | |
125 $ get-with-headers.py --headeronly 127.0.0.1:$HGPORT '?cmd=getbundle&heads=e93700bd72895c5addab234c56d4024b487a362f&common=0000000000000000000000000000000000000000' Content-Type | |
126 200 Script output follows | |
127 content-type: application/mercurial-0.1 | |
128 | |
129 $ get-with-headers.py 127.0.0.1:$HGPORT '?cmd=getbundle&heads=e93700bd72895c5addab234c56d4024b487a362f&common=0000000000000000000000000000000000000000' > resp | |
130 $ f --size --hexdump --bytes 28 --sha1 resp | |
131 resp: size=227, sha1=35a4c074da74f32f5440da3cbf04 | |
132 0000: 32 30 30 20 53 63 72 69 70 74 20 6f 75 74 70 75 |200 Script outpu| | |
133 0010: 74 20 66 6f 6c 6c 6f 77 73 0a 0a 78 |t follows..x| | |
134 | |
135 Explicit 0.1 will send zlib because "none" isn't supported on 0.1 | |
136 | |
137 $ get-with-headers.py --hgproto '0.1' 127.0.0.1:$HGPORT '?cmd=getbundle&heads=e93700bd72895c5addab234c56d4024b487a362f&common=0000000000000000000000000000000000000000' > resp | |
138 $ f --size --hexdump --bytes 28 --sha1 resp | |
139 resp: size=227, sha1=35a4c074da74f32f5440da3cbf04 | |
140 0000: 32 30 30 20 53 63 72 69 70 74 20 6f 75 74 70 75 |200 Script outpu| | |
141 0010: 74 20 66 6f 6c 6c 6f 77 73 0a 0a 78 |t follows..x| | |
142 | |
143 0.2 with no compression will get "none" because that is server's preference | |
144 (spec says ZL and UN are implicitly supported) | |
145 | |
146 $ get-with-headers.py --hgproto '0.2' 127.0.0.1:$HGPORT '?cmd=getbundle&heads=e93700bd72895c5addab234c56d4024b487a362f&common=0000000000000000000000000000000000000000' > resp | |
147 $ f --size --hexdump --bytes 32 --sha1 resp | |
148 resp: size=432, sha1=ac931b412ec185a02e0e5bcff98dac83 | |
149 0000: 32 30 30 20 53 63 72 69 70 74 20 6f 75 74 70 75 |200 Script outpu| | |
150 0010: 74 20 66 6f 6c 6c 6f 77 73 0a 0a 04 6e 6f 6e 65 |t follows...none| | |
151 | |
152 Client receives server preference even if local order doesn't match | |
153 | |
154 $ get-with-headers.py --hgproto '0.2 comp=zlib,none' 127.0.0.1:$HGPORT '?cmd=getbundle&heads=e93700bd72895c5addab234c56d4024b487a362f&common=0000000000000000000000000000000000000000' > resp | |
155 $ f --size --hexdump --bytes 32 --sha1 resp | |
156 resp: size=432, sha1=ac931b412ec185a02e0e5bcff98dac83 | |
157 0000: 32 30 30 20 53 63 72 69 70 74 20 6f 75 74 70 75 |200 Script outpu| | |
158 0010: 74 20 66 6f 6c 6c 6f 77 73 0a 0a 04 6e 6f 6e 65 |t follows...none| | |
159 | |
160 Client receives only supported format even if not server preferred format | |
161 | |
162 $ get-with-headers.py --hgproto '0.2 comp=zlib' 127.0.0.1:$HGPORT '?cmd=getbundle&heads=e93700bd72895c5addab234c56d4024b487a362f&common=0000000000000000000000000000000000000000' > resp | |
163 $ f --size --hexdump --bytes 33 --sha1 resp | |
164 resp: size=232, sha1=a1c727f0c9693ca15742a75c30419bc36 | |
165 0000: 32 30 30 20 53 63 72 69 70 74 20 6f 75 74 70 75 |200 Script outpu| | |
166 0010: 74 20 66 6f 6c 6c 6f 77 73 0a 0a 04 7a 6c 69 62 |t follows...zlib| | |
167 0020: 78 |x| |