@@ -5,6 +5,7 @@
import logging
import socket
+import asyncio
import bb.asyncrpc
import json
from . import create_async_client
@@ -13,6 +14,63 @@ from . import create_async_client
logger = logging.getLogger("hashserv.client")
+class Batch(object):
+ def __init__(self):
+ self.done = False
+ self.cond = asyncio.Condition()
+ self.pending = []
+ self.results = []
+ self.sent_count = 0
+
+ async def recv(self, socket):
+ while True:
+ async with self.cond:
+ await self.cond.wait_for(lambda: self.pending or self.done)
+
+ if not self.pending:
+ if self.done:
+ return
+ continue
+
+ r = await socket.recv()
+ self.results.append(r)
+
+ async with self.cond:
+ self.pending.pop(0)
+
+ async def send(self, socket, msgs):
+ try:
+ # In the event of a restart due to a reconnect, all in-flight
+ # messages need to be resent first to keep to result count in sync
+ for m in self.pending:
+ await socket.send(m)
+
+ for m in msgs:
+ await socket.send(m)
+
+ async with self.cond:
+ self.pending.append(m)
+ self.cond.notify()
+ self.sent_count += 1
+ finally:
+ async with self.cond:
+ self.done = True
+ self.cond.notify()
+
+ async def process(self, socket, msgs):
+ await asyncio.gather(
+ self.recv(socket),
+ self.send(socket, msgs),
+ )
+
+ if len(self.results) != self.sent_count:
+ raise ValueError(
+ f"Expected result count {len(self.results)}. Expected {self.sent_count}"
+ )
+
+ return self.results
+
+
class AsyncClient(bb.asyncrpc.AsyncClient):
MODE_NORMAL = 0
MODE_GET_STREAM = 1
@@ -36,11 +94,27 @@ class AsyncClient(bb.asyncrpc.AsyncClient):
if become:
await self.become_user(become)
- async def send_stream(self, mode, msg):
+ async def send_stream_batch(self, mode, msgs):
+ """
+ Does a "batch" process of stream messages. This sends the query
+ messages as fast as possible, and simultaneously attempts to read the
+ messages back. This helps to mitigate the effects of latency to the
+ hash equivalence server be allowing multiple queries to be "in-flight"
+ at once
+
+ The implementation does more complicated tracking using a count of sent
+ messages so that `msgs` can be a generator function (i.e. its length is
+ unknown)
+
+ """
+
+ b = Batch()
+
async def proc():
+ nonlocal b
+
await self._set_mode(mode)
- await self.socket.send(msg)
- return await self.socket.recv()
+ return await b.process(self.socket, msgs)
return await self._send_wrapper(proc)
@@ -89,10 +163,15 @@ class AsyncClient(bb.asyncrpc.AsyncClient):
self.mode = new_mode
async def get_unihash(self, method, taskhash):
- r = await self.send_stream(self.MODE_GET_STREAM, "%s %s" % (method, taskhash))
- if not r:
- return None
- return r
+ r = await self.get_unihash_batch([(method, taskhash)])
+ return r[0]
+
+ async def get_unihash_batch(self, args):
+ result = await self.send_stream_batch(
+ self.MODE_GET_STREAM,
+ (f"{method} {taskhash}" for method, taskhash in args),
+ )
+ return [r if r else None for r in result]
async def report_unihash(self, taskhash, method, outhash, unihash, extra={}):
m = extra.copy()
@@ -115,8 +194,12 @@ class AsyncClient(bb.asyncrpc.AsyncClient):
)
async def unihash_exists(self, unihash):
- r = await self.send_stream(self.MODE_EXIST_STREAM, unihash)
- return r == "true"
+ r = await self.unihash_exists_batch([unihash])
+ return r[0]
+
+ async def unihash_exists_batch(self, unihashes):
+ result = await self.send_stream_batch(self.MODE_EXIST_STREAM, unihashes)
+ return [r == "true" for r in result]
async def get_outhash(self, method, outhash, taskhash, with_unihash=True):
return await self.invoke(
@@ -237,10 +320,12 @@ class Client(bb.asyncrpc.Client):
"connect_tcp",
"connect_websocket",
"get_unihash",
+ "get_unihash_batch",
"report_unihash",
"report_unihash_equiv",
"get_taskhash",
"unihash_exists",
+ "unihash_exists_batch",
"get_outhash",
"get_stats",
"reset_stats",
@@ -594,6 +594,43 @@ class HashEquivalenceCommonTests(object):
7: None,
})
+ def test_get_unihash_batch(self):
+ TEST_INPUT = (
+ # taskhash outhash unihash
+ ('8aa96fcffb5831b3c2c0cb75f0431e3f8b20554a', 'afe240a439959ce86f5e322f8c208e1fedefea9e813f2140c81af866cc9edf7e','218e57509998197d570e2c98512d0105985dffc9'),
+ # Duplicated taskhash with multiple output hashes and unihashes.
+ ('8aa96fcffb5831b3c2c0cb75f0431e3f8b20554a', '0904a7fe3dc712d9fd8a74a616ddca2a825a8ee97adf0bd3fc86082c7639914d', 'ae9a7d252735f0dafcdb10e2e02561ca3a47314c'),
+ # Equivalent hash
+ ("044c2ec8aaf480685a00ff6ff49e6162e6ad34e1", '0904a7fe3dc712d9fd8a74a616ddca2a825a8ee97adf0bd3fc86082c7639914d', "def64766090d28f627e816454ed46894bb3aab36"),
+ ("e3da00593d6a7fb435c7e2114976c59c5fd6d561", "1cf8713e645f491eb9c959d20b5cae1c47133a292626dda9b10709857cbe688a", "3b5d3d83f07f259e9086fcb422c855286e18a57d"),
+ ('35788efcb8dfb0a02659d81cf2bfd695fb30faf9', '2765d4a5884be49b28601445c2760c5f21e7e5c0ee2b7e3fce98fd7e5970796f', 'f46d3fbb439bd9b921095da657a4de906510d2cd'),
+ ('35788efcb8dfb0a02659d81cf2bfd695fb30fafa', '2765d4a5884be49b28601445c2760c5f21e7e5c0ee2b7e3fce98fd7e5970796f', 'f46d3fbb439bd9b921095da657a4de906510d2ce'),
+ ('9d81d76242cc7cfaf7bf74b94b9cd2e29324ed74', '8470d56547eea6236d7c81a644ce74670ca0bbda998e13c629ef6bb3f0d60b69', '05d2a63c81e32f0a36542ca677e8ad852365c538'),
+ )
+ EXTRA_QUERIES = (
+ "6b6be7a84ab179b4240c4302518dc3f6",
+ )
+
+ for taskhash, outhash, unihash in TEST_INPUT:
+ self.client.report_unihash(taskhash, self.METHOD, outhash, unihash)
+
+
+ result = self.client.get_unihash_batch(
+ [(self.METHOD, data[0]) for data in TEST_INPUT] +
+ [(self.METHOD, e) for e in EXTRA_QUERIES]
+ )
+
+ self.assertListEqual(result, [
+ "218e57509998197d570e2c98512d0105985dffc9",
+ "218e57509998197d570e2c98512d0105985dffc9",
+ "218e57509998197d570e2c98512d0105985dffc9",
+ "3b5d3d83f07f259e9086fcb422c855286e18a57d",
+ "f46d3fbb439bd9b921095da657a4de906510d2cd",
+ "f46d3fbb439bd9b921095da657a4de906510d2cd",
+ "05d2a63c81e32f0a36542ca677e8ad852365c538",
+ None,
+ ])
+
def test_client_pool_unihash_exists(self):
TEST_INPUT = (
# taskhash outhash unihash
@@ -636,6 +673,44 @@ class HashEquivalenceCommonTests(object):
result = client_pool.unihashes_exist(query)
self.assertDictEqual(result, expected)
+ def test_unihash_exists_batch(self):
+ TEST_INPUT = (
+ # taskhash outhash unihash
+ ('8aa96fcffb5831b3c2c0cb75f0431e3f8b20554a', 'afe240a439959ce86f5e322f8c208e1fedefea9e813f2140c81af866cc9edf7e','218e57509998197d570e2c98512d0105985dffc9'),
+ # Duplicated taskhash with multiple output hashes and unihashes.
+ ('8aa96fcffb5831b3c2c0cb75f0431e3f8b20554a', '0904a7fe3dc712d9fd8a74a616ddca2a825a8ee97adf0bd3fc86082c7639914d', 'ae9a7d252735f0dafcdb10e2e02561ca3a47314c'),
+ # Equivalent hash
+ ("044c2ec8aaf480685a00ff6ff49e6162e6ad34e1", '0904a7fe3dc712d9fd8a74a616ddca2a825a8ee97adf0bd3fc86082c7639914d', "def64766090d28f627e816454ed46894bb3aab36"),
+ ("e3da00593d6a7fb435c7e2114976c59c5fd6d561", "1cf8713e645f491eb9c959d20b5cae1c47133a292626dda9b10709857cbe688a", "3b5d3d83f07f259e9086fcb422c855286e18a57d"),
+ ('35788efcb8dfb0a02659d81cf2bfd695fb30faf9', '2765d4a5884be49b28601445c2760c5f21e7e5c0ee2b7e3fce98fd7e5970796f', 'f46d3fbb439bd9b921095da657a4de906510d2cd'),
+ ('35788efcb8dfb0a02659d81cf2bfd695fb30fafa', '2765d4a5884be49b28601445c2760c5f21e7e5c0ee2b7e3fce98fd7e5970796f', 'f46d3fbb439bd9b921095da657a4de906510d2ce'),
+ ('9d81d76242cc7cfaf7bf74b94b9cd2e29324ed74', '8470d56547eea6236d7c81a644ce74670ca0bbda998e13c629ef6bb3f0d60b69', '05d2a63c81e32f0a36542ca677e8ad852365c538'),
+ )
+ EXTRA_QUERIES = (
+ "6b6be7a84ab179b4240c4302518dc3f6",
+ )
+
+ result_unihashes = set()
+
+
+ for taskhash, outhash, unihash in TEST_INPUT:
+ result = self.client.report_unihash(taskhash, self.METHOD, outhash, unihash)
+ result_unihashes.add(result["unihash"])
+
+ query = []
+ expected = []
+
+ for _, _, unihash in TEST_INPUT:
+ query.append(unihash)
+ expected.append(unihash in result_unihashes)
+
+
+ for unihash in EXTRA_QUERIES:
+ query.append(unihash)
+ expected.append(False)
+
+ result = self.client.unihash_exists_batch(query)
+ self.assertListEqual(result, expected)
def test_auth_read_perms(self):
admin_client = self.start_auth_server()
Changes the stream mode to do "batch" processing. This means that the sending and reciving of messages is done simultaneously so that messages can be sent as fast as possible without having to wait for each reply. This allows multiple messages to be in flight at once, reducing the effect of the round trip latency from the server. Signed-off-by: Joshua Watt <JPEWhacker@gmail.com> --- bitbake/lib/hashserv/client.py | 103 ++++++++++++++++++++++++++++++--- bitbake/lib/hashserv/tests.py | 75 ++++++++++++++++++++++++ 2 files changed, 169 insertions(+), 9 deletions(-)