Ticket #1392: testgetlatencies01.darcs.patch

Fri Mar 25 14:35:14 MDT 2011  wilcoxjg@gmail.com
  * storage: new mocking tests of storage server read and write
  There are already tests of read and functionality in test_storage.py, but those tests let the code under test use a real filesystem whereas these tests mock all file system calls.

Thu Apr 14 16:48:23 MDT 2011  zooko@zooko.com
  * test_server.py --> test_backends.py:  server.py: added testing of get_latencies in StorageServer
  This patch test both coverage and handling of small samples in the get_latencies method of StorageServer.  get_latencies now distinguishes between highly repetitive latencies and small sample sizes.  This is of most concern at the big end of the latency distribution, although the ambiguity increases in general as the sample size decreases.

Thu Apr 14 17:15:26 MDT 2011  zooko@zooko.com
  * test_backends: cleaned whitespace, made test_get_latencies it's own function

New patches:

[storage: new mocking tests of storage server read and write
wilcoxjg@gmail.com**20110325203514
 Ignore-this: df65c3c4f061dd1516f88662023fdb41
 There are already tests of read and functionality in test_storage.py, but those tests let the code under test use a real filesystem whereas these tests mock all file system calls.
] {
addfile ./src/allmydata/test/test_server.py
hunk ./src/allmydata/test/test_server.py 1
+from twisted.trial import unittest
+
+from StringIO import StringIO
+
+from allmydata.test.common_util import ReallyEqualMixin
+
+import mock
+
+# This is the code that we're going to be testing.
+from allmydata.storage.server import StorageServer
+
+# The following share file contents was generated with
+# storage.immutable.ShareFile from Tahoe-LAFS v1.8.2
+# with share data == 'a'.
+share_data = 'a\x00\x00\x00\x00xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyy\x00(\xde\x80'
+share_file_data = '\x00\x00\x00\x01\x00\x00\x00\x01\x00\x00\x00\x01' + share_data
+
+sharefname = 'testdir/shares/or/orsxg5dtorxxeylhmvpws3temv4a/0'
+
+class TestServerConstruction(unittest.TestCase, ReallyEqualMixin):
+    @mock.patch('__builtin__.open')
+    def test_create_server(self, mockopen):
+        """ This tests whether a server instance can be constructed. """
+
+        def call_open(fname, mode):
+            if fname == 'testdir/bucket_counter.state':
+                raise IOError(2, "No such file or directory: 'testdir/bucket_counter.state'")
+            elif fname == 'testdir/lease_checker.state':
+                raise IOError(2, "No such file or directory: 'testdir/lease_checker.state'")
+            elif fname == 'testdir/lease_checker.history':
+                return StringIO()
+        mockopen.side_effect = call_open
+
+        # Now begin the test.
+        s = StorageServer('testdir', 'testnodeidxxxxxxxxxx')
+
+        # You passed!
+
+class TestServer(unittest.TestCase, ReallyEqualMixin):
+    @mock.patch('__builtin__.open')
+    def setUp(self, mockopen):
+        def call_open(fname, mode):
+            if fname == 'testdir/bucket_counter.state':
+                raise IOError(2, "No such file or directory: 'testdir/bucket_counter.state'")
+            elif fname == 'testdir/lease_checker.state':
+                raise IOError(2, "No such file or directory: 'testdir/lease_checker.state'")
+            elif fname == 'testdir/lease_checker.history':
+                return StringIO()
+        mockopen.side_effect = call_open
+
+        self.s = StorageServer('testdir', 'testnodeidxxxxxxxxxx')
+
+
+    @mock.patch('time.time')
+    @mock.patch('os.mkdir')
+    @mock.patch('__builtin__.open')
+    @mock.patch('os.listdir')
+    @mock.patch('os.path.isdir')
+    def test_write_share(self, mockisdir, mocklistdir, mockopen, mockmkdir, mocktime):
+        """Handle a report of corruption."""
+
+        def call_listdir(dirname):
+            self.failUnlessReallyEqual(dirname, 'testdir/shares/or/orsxg5dtorxxeylhmvpws3temv4a')
+            raise OSError(2, "No such file or directory: 'testdir/shares/or/orsxg5dtorxxeylhmvpws3temv4a'")
+
+        mocklistdir.side_effect = call_listdir
+
+        class MockFile:
+            def __init__(self):
+                self.buffer = ''
+                self.pos = 0
+            def write(self, instring):
+                begin = self.pos
+                padlen = begin - len(self.buffer)
+                if padlen > 0:
+                    self.buffer += '\x00' * padlen
+                end = self.pos + len(instring)
+                self.buffer = self.buffer[:begin]+instring+self.buffer[end:]
+                self.pos = end
+            def close(self):
+                pass
+            def seek(self, pos):
+                self.pos = pos
+            def read(self, numberbytes):
+                return self.buffer[self.pos:self.pos+numberbytes]
+            def tell(self):
+                return self.pos
+
+        mocktime.return_value = 0
+
+        sharefile = MockFile()
+        def call_open(fname, mode):
+            self.failUnlessReallyEqual(fname, 'testdir/shares/incoming/or/orsxg5dtorxxeylhmvpws3temv4a/0' )
+            return sharefile
+
+        mockopen.side_effect = call_open
+        # Now begin the test.
+        alreadygot, bs = self.s.remote_allocate_buckets('teststorage_index', 'x'*32, 'y'*32, set((0,)), 1, mock.Mock())
+        print bs
+        bs[0].remote_write(0, 'a')
+        self.failUnlessReallyEqual(sharefile.buffer, share_file_data)
+
+
+    @mock.patch('os.path.exists')
+    @mock.patch('os.path.getsize')
+    @mock.patch('__builtin__.open')
+    @mock.patch('os.listdir')
+    def test_read_share(self, mocklistdir, mockopen, mockgetsize, mockexists):
+        """ This tests whether the code correctly finds and reads
+        shares written out by old (Tahoe-LAFS <= v1.8.2)
+        servers. There is a similar test in test_download, but that one
+        is from the perspective of the client and exercises a deeper
+        stack of code. This one is for exercising just the
+        StorageServer object. """
+
+        def call_listdir(dirname):
+            self.failUnlessReallyEqual(dirname,'testdir/shares/or/orsxg5dtorxxeylhmvpws3temv4a')
+            return ['0']
+
+        mocklistdir.side_effect = call_listdir
+
+        def call_open(fname, mode):
+            self.failUnlessReallyEqual(fname, sharefname)
+            self.failUnless('r' in mode, mode)
+            self.failUnless('b' in mode, mode)
+
+            return StringIO(share_file_data)
+        mockopen.side_effect = call_open
+
+        datalen = len(share_file_data)
+        def call_getsize(fname):
+            self.failUnlessReallyEqual(fname, sharefname)
+            return datalen
+        mockgetsize.side_effect = call_getsize
+
+        def call_exists(fname):
+            self.failUnlessReallyEqual(fname, sharefname)
+            return True
+        mockexists.side_effect = call_exists
+
+        # Now begin the test.
+        bs = self.s.remote_get_buckets('teststorage_index')
+
+        self.failUnlessEqual(len(bs), 1)
+        b = bs[0]
+        self.failUnlessReallyEqual(b.remote_read(0, datalen), share_data)
+        # If you try to read past the end you get the as much data as is there.
+        self.failUnlessReallyEqual(b.remote_read(0, datalen+20), share_data)
+        # If you start reading past the end of the file you get the empty string.
+        self.failUnlessReallyEqual(b.remote_read(datalen+1, 3), '')
}
[test_server.py --> test_backends.py:  server.py: added testing of get_latencies in StorageServer
zooko@zooko.com**20110414224823
 Ignore-this: 3e266de570f725f768d18c131e2c6d8
 This patch test both coverage and handling of small samples in the get_latencies method of StorageServer.  get_latencies now distinguishes between highly repetitive latencies and small sample sizes.  This is of most concern at the big end of the latency distribution, although the ambiguity increases in general as the sample size decreases.
] {
move ./src/allmydata/test/test_server.py ./src/allmydata/test/test_backends.py
hunk ./src/allmydata/storage/server.py 134
             samples = self.latencies[category][:]
             samples.sort()
             count = len(samples)
+            if count < 1000:
+                output[category] = None
+                continue
+            samples.sort()
+            stats = {}
             stats["mean"] = sum(samples) / count
             stats["01_0_percentile"] = samples[int(0.01 * count)]
             stats["10_0_percentile"] = samples[int(0.1 * count)]
hunk ./src/allmydata/test/test_backends.py 21
 sharefname = 'testdir/shares/or/orsxg5dtorxxeylhmvpws3temv4a/0'
 
 class TestServerConstruction(unittest.TestCase, ReallyEqualMixin):
+    @mock.patch('time.time')
+    @mock.patch('os.mkdir')
+    @mock.patch('__builtin__.open')
+    @mock.patch('os.listdir')
+    @mock.patch('os.path.isdir')
+    def test_create_server_null_backend(self, mockisdir, mocklistdir, mockopen, mockmkdir, mocktime):
+        """ This tests whether a server instance can be constructed
+        with a null backend. The server instance fails the test if it
+        tries to read or write to the file system. """
+
+        # Now begin the test.
+        s = StorageServer('testnodeidxxxxxxxxxx', backend=NullBackend(),stats_provider=mock.Mock())
+
+        # The null backend should not talk to the os.
+        self.failIf(mockisdir.called)
+        self.failIf(mocklistdir.called)
+        self.failIf(mockopen.called)
+        self.failIf(mockmkdir.called)
+        #self.failIf(mocktime.called)
+        
+        #  The server's representation should not change.
+        self.failUnlessReallyEqual(s.__repr__(),'<StorageServer orsxg5do>')
+
+        #  There should be no latencies when the backend is null.  *** The "cancel" category is left out to increase coverage.  This seems like a dubious decision.  Must consult with more knowledgeable persons.
+        numbersamples = 1001
+        for category in ["allocate","write","close","read","get",\
+                         "writev","readv","add-lease","renew"]:#,"cancel"]:
+            [s.add_latency(category,x) for x in numbersamples*[0]]
+        l = s.get_latencies()
+
+        # Now test that get_latencies correctly reports None for small sample-sizes.
+        s1 = StorageServer('testnodeidxxxxxxxxxx', backend=NullBackend(),stats_provider=mock.Mock())
+        #  There should be no latencies when the backend is null.  *** The "cancel" category is left out to increase coverage.  This seems like a dubious decision.  Must consult with more knowledgeable persons.
+        numbersamples = 10
+        for category in ["allocate","write","close","read","get",\
+                         "writev","readv","add-lease","renew"]:#,"cancel"]:
+            [s1.add_latency(category,x) for x in numbersamples*[0]]
+        l1 = s1.get_latencies()
+
+        for key in l1.keys():
+            self.failUnlessReallyEqual(l1[key],None)
+        # You passed!
+
+    @mock.patch('time.time')
+    @mock.patch('os.mkdir')
     @mock.patch('__builtin__.open')
     def test_create_server(self, mockopen):
         """ This tests whether a server instance can be constructed. """
}
[test_backends: cleaned whitespace, made test_get_latencies it's own function
zooko@zooko.com**20110414231526
 Ignore-this: 483e7d4eb67cc0cbf780d8219156907e
] {
hunk ./src/allmydata/test/test_backends.py 32
         tries to read or write to the file system. """
 
         # Now begin the test.
-        s = StorageServer('testnodeidxxxxxxxxxx', backend=NullBackend(),stats_provider=mock.Mock())
+        s = StorageServer('testnodeidxxxxxxxxxx', backend=NullBackend(), stats_provider=mock.Mock())
 
         # The null backend should not talk to the os.
         self.failIf(mockisdir.called)
hunk ./src/allmydata/test/test_backends.py 40
         self.failIf(mockopen.called)
         self.failIf(mockmkdir.called)
         #self.failIf(mocktime.called)
-        
+
         #  The server's representation should not change.
hunk ./src/allmydata/test/test_backends.py 42
-        self.failUnlessReallyEqual(s.__repr__(),'<StorageServer orsxg5do>')
+        self.failUnlessReallyEqual(s.__repr__(), '<StorageServer orsxg5do>')
+
+        # You passed!
+
+    @mock.patch('time.time')
+    @mock.patch('os.mkdir')
+    @mock.patch('__builtin__.open')
+    @mock.patch('os.listdir')
+    @mock.patch('os.path.isdir')
+    def test_get_latencies(self, mockisdir, mocklistdir, mockopen, mockmkdir, mocktime):
+        """  There should be no latencies when the backend is null.  *** The "cancel" category is left out to increase coverage.  This seems like a dubious decision.  Must consult with more knowledgeable persons."""
+
+        # Now begin the test.
+        s = StorageServer('testnodeidxxxxxxxxxx', backend=NullBackend(), stats_provider=mock.Mock())
+
+        # The null backend should not talk to the os.
+        self.failIf(mockisdir.called)
+        self.failIf(mocklistdir.called)
+        self.failIf(mockopen.called)
+        self.failIf(mockmkdir.called)
+        #self.failIf(mocktime.called)
 
hunk ./src/allmydata/test/test_backends.py 64
-        #  There should be no latencies when the backend is null.  *** The "cancel" category is left out to increase coverage.  This seems like a dubious decision.  Must consult with more knowledgeable persons.
         numbersamples = 1001
hunk ./src/allmydata/test/test_backends.py 65
-        for category in ["allocate","write","close","read","get",\
-                         "writev","readv","add-lease","renew"]:#,"cancel"]:
-            [s.add_latency(category,x) for x in numbersamples*[0]]
+        for category in ["allocate", "write", "close", "read", "get", \
+                         "writev", "readv", "add-lease", "renew"]:#,"cancel"]:
+            [s.add_latency(category, x) for x in numbersamples*[0]]
         l = s.get_latencies()
 
         # Now test that get_latencies correctly reports None for small sample-sizes.
hunk ./src/allmydata/test/test_backends.py 71
-        s1 = StorageServer('testnodeidxxxxxxxxxx', backend=NullBackend(),stats_provider=mock.Mock())
+        s1 = StorageServer('testnodeidxxxxxxxxxx', backend=NullBackend(), stats_provider=mock.Mock())
         #  There should be no latencies when the backend is null.  *** The "cancel" category is left out to increase coverage.  This seems like a dubious decision.  Must consult with more knowledgeable persons.
         numbersamples = 10
hunk ./src/allmydata/test/test_backends.py 74
-        for category in ["allocate","write","close","read","get",\
-                         "writev","readv","add-lease","renew"]:#,"cancel"]:
-            [s1.add_latency(category,x) for x in numbersamples*[0]]
+        for category in ["allocate", "write", "close", "read", "get", \
+                         "writev", "readv", "add-lease", "renew"]:#, "cancel"]:
+            [s1.add_latency(category, x) for x in numbersamples*[0]]
         l1 = s1.get_latencies()
 
         for key in l1.keys():
hunk ./src/allmydata/test/test_backends.py 80
-            self.failUnlessReallyEqual(l1[key],None)
+            self.failUnlessReallyEqual(l1[key], None)
         # You passed!
 
     @mock.patch('time.time')
hunk ./src/allmydata/test/test_backends.py 167
         bs[0].remote_write(0, 'a')
         self.failUnlessReallyEqual(sharefile.buffer, share_file_data)
 
-
     @mock.patch('os.path.exists')
     @mock.patch('os.path.getsize')
     @mock.patch('__builtin__.open')
}

Context:

[Fix a test failure in test_package_initialization on Python 2.4.x due to exceptions being stringified differently than in later versions of Python. refs #1389
david-sarah@jacaranda.org**20110411190738
 Ignore-this: 7847d26bc117c328c679f08a7baee519
] 
[tests: add test for including the ImportError message and traceback entry in the summary of errors from importing dependencies. refs #1389
david-sarah@jacaranda.org**20110410155844
 Ignore-this: fbecdbeb0d06a0f875fe8d4030aabafa
] 
[allmydata/__init__.py: preserve the message and last traceback entry (file, line number, function, and source line) of ImportErrors in the package versions string. fixes #1389
david-sarah@jacaranda.org**20110410155705
 Ignore-this: 2f87b8b327906cf8bfca9440a0904900
] 
[remove unused variable detected by pyflakes
zooko@zooko.com**20110407172231
 Ignore-this: 7344652d5e0720af822070d91f03daf9
] 
[allmydata/__init__.py: Nicer reporting of unparseable version numbers in dependencies. fixes #1388
david-sarah@jacaranda.org**20110401202750
 Ignore-this: 9c6bd599259d2405e1caadbb3e0d8c7f
] 
[update FTP-and-SFTP.rst: the necessary patch is included in Twisted-10.1
Brian Warner <warner@lothar.com>**20110325232511
 Ignore-this: d5307faa6900f143193bfbe14e0f01a
] 
[control.py: remove all uses of s.get_serverid()
warner@lothar.com**20110227011203
 Ignore-this: f80a787953bd7fa3d40e828bde00e855
] 
[web: remove some uses of s.get_serverid(), not all
warner@lothar.com**20110227011159
 Ignore-this: a9347d9cf6436537a47edc6efde9f8be
] 
[immutable/downloader/fetcher.py: remove all get_serverid() calls
warner@lothar.com**20110227011156
 Ignore-this: fb5ef018ade1749348b546ec24f7f09a
] 
[immutable/downloader/fetcher.py: fix diversity bug in server-response handling
warner@lothar.com**20110227011153
 Ignore-this: bcd62232c9159371ae8a16ff63d22c1b
 
 When blocks terminate (either COMPLETE or CORRUPT/DEAD/BADSEGNUM), the
 _shares_from_server dict was being popped incorrectly (using shnum as the
 index instead of serverid). I'm still thinking through the consequences of
 this bug. It was probably benign and really hard to detect. I think it would
 cause us to incorrectly believe that we're pulling too many shares from a
 server, and thus prefer a different server rather than asking for a second
 share from the first server. The diversity code is intended to spread out the
 number of shares simultaneously being requested from each server, but with
 this bug, it might be spreading out the total number of shares requested at
 all, not just simultaneously. (note that SegmentFetcher is scoped to a single
 segment, so the effect doesn't last very long).
] 
[immutable/downloader/share.py: reduce get_serverid(), one left, update ext deps
warner@lothar.com**20110227011150
 Ignore-this: d8d56dd8e7b280792b40105e13664554
 
 test_download.py: create+check MyShare instances better, make sure they share
 Server objects, now that finder.py cares
] 
[immutable/downloader/finder.py: reduce use of get_serverid(), one left
warner@lothar.com**20110227011146
 Ignore-this: 5785be173b491ae8a78faf5142892020
] 
[immutable/offloaded.py: reduce use of get_serverid() a bit more
warner@lothar.com**20110227011142
 Ignore-this: b48acc1b2ae1b311da7f3ba4ffba38f
] 
[immutable/upload.py: reduce use of get_serverid()
warner@lothar.com**20110227011138
 Ignore-this: ffdd7ff32bca890782119a6e9f1495f6
] 
[immutable/checker.py: remove some uses of s.get_serverid(), not all
warner@lothar.com**20110227011134
 Ignore-this: e480a37efa9e94e8016d826c492f626e
] 
[add remaining get_* methods to storage_client.Server, NoNetworkServer, and
warner@lothar.com**20110227011132
 Ignore-this: 6078279ddf42b179996a4b53bee8c421
 MockIServer stubs
] 
[upload.py: rearrange _make_trackers a bit, no behavior changes
warner@lothar.com**20110227011128
 Ignore-this: 296d4819e2af452b107177aef6ebb40f
] 
[happinessutil.py: finally rename merge_peers to merge_servers
warner@lothar.com**20110227011124
 Ignore-this: c8cd381fea1dd888899cb71e4f86de6e
] 
[test_upload.py: factor out FakeServerTracker
warner@lothar.com**20110227011120
 Ignore-this: 6c182cba90e908221099472cc159325b
] 
[test_upload.py: server-vs-tracker cleanup
warner@lothar.com**20110227011115
 Ignore-this: 2915133be1a3ba456e8603885437e03
] 
[happinessutil.py: server-vs-tracker cleanup
warner@lothar.com**20110227011111
 Ignore-this: b856c84033562d7d718cae7cb01085a9
] 
[upload.py: more tracker-vs-server cleanup
warner@lothar.com**20110227011107
 Ignore-this: bb75ed2afef55e47c085b35def2de315
] 
[upload.py: fix var names to avoid confusion between 'trackers' and 'servers'
warner@lothar.com**20110227011103
 Ignore-this: 5d5e3415b7d2732d92f42413c25d205d
] 
[refactor: s/peer/server/ in immutable/upload, happinessutil.py, test_upload
warner@lothar.com**20110227011100
 Ignore-this: 7ea858755cbe5896ac212a925840fe68
 
 No behavioral changes, just updating variable/method names and log messages.
 The effects outside these three files should be minimal: some exception
 messages changed (to say "server" instead of "peer"), and some internal class
 names were changed. A few things still use "peer" to minimize external
 changes, like UploadResults.timings["peer_selection"] and
 happinessutil.merge_peers, which can be changed later.
] 
[storage_client.py: clean up test_add_server/test_add_descriptor, remove .test_servers
warner@lothar.com**20110227011056
 Ignore-this: efad933e78179d3d5fdcd6d1ef2b19cc
] 
[test_client.py, upload.py:: remove KiB/MiB/etc constants, and other dead code
warner@lothar.com**20110227011051
 Ignore-this: dc83c5794c2afc4f81e592f689c0dc2d
] 
[test: increase timeout on a network test because Francois's ARM machine hit that timeout
zooko@zooko.com**20110317165909
 Ignore-this: 380c345cdcbd196268ca5b65664ac85b
 I'm skeptical that the test was proceeding correctly but ran out of time. It seems more likely that it had gotten hung. But if we raise the timeout to an even more extravagant number then we can be even more certain that the test was never going to finish.
] 
[docs/configuration.rst: add a "Frontend Configuration" section
Brian Warner <warner@lothar.com>**20110222014323
 Ignore-this: 657018aa501fe4f0efef9851628444ca
 
 this points to docs/frontends/*.rst, which were previously underlinked
] 
[web/filenode.py: avoid calling req.finish() on closed HTTP connections. Closes #1366
"Brian Warner <warner@lothar.com>"**20110221061544
 Ignore-this: 799d4de19933f2309b3c0c19a63bb888
] 
[Add unit tests for cross_check_pkg_resources_versus_import, and a regression test for ref #1355. This requires a little refactoring to make it testable.
david-sarah@jacaranda.org**20110221015817
 Ignore-this: 51d181698f8c20d3aca58b057e9c475a
] 
[allmydata/__init__.py: .name was used in place of the correct .__name__ when printing an exception. Also, robustify string formatting by using %r instead of %s in some places. fixes #1355.
david-sarah@jacaranda.org**20110221020125
 Ignore-this: b0744ed58f161bf188e037bad077fc48
] 
[Refactor StorageFarmBroker handling of servers
Brian Warner <warner@lothar.com>**20110221015804
 Ignore-this: 842144ed92f5717699b8f580eab32a51
 
 Pass around IServer instance instead of (peerid, rref) tuple. Replace
 "descriptor" with "server". Other replacements:
 
  get_all_servers -> get_connected_servers/get_known_servers
  get_servers_for_index -> get_servers_for_psi (now returns IServers)
 
 This change still needs to be pushed further down: lots of code is now
 getting the IServer and then distributing (peerid, rref) internally.
 Instead, it ought to distribute the IServer internally and delay
 extracting a serverid or rref until the last moment.
 
 no_network.py was updated to retain parallelism.
] 
[TAG allmydata-tahoe-1.8.2
warner@lothar.com**20110131020101] 
Patch bundle hash:
48f5cedd889bc9d05be6343375b5908a8c42d000