Ticket #999: workingonbackend01.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.

Fri Jun 24 14:28:50 MDT 2011  wilcoxjg@gmail.com
  * server.py, test_backends.py, interfaces.py, immutable.py (others?): working patch for implementation of backends plugin
  sloppy not for production

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), '')
}
[server.py, test_backends.py, interfaces.py, immutable.py (others?): working patch for implementation of backends plugin
wilcoxjg@gmail.com**20110624202850
 Ignore-this: ca6f34987ee3b0d25cac17c1fc22d50c
 sloppy not for production
] {
move ./src/allmydata/test/test_server.py ./src/allmydata/test/test_backends.py
hunk ./src/allmydata/storage/crawler.py 13
     pass
 
 class ShareCrawler(service.MultiService):
-    """A ShareCrawler subclass is attached to a StorageServer, and
+    """A subcless of ShareCrawler is attached to a StorageServer, and
     periodically walks all of its shares, processing each one in some
     fashion. This crawl is rate-limited, to reduce the IO burden on the host,
     since large servers can easily have a terabyte of shares, in several
hunk ./src/allmydata/storage/crawler.py 31
     We assume that the normal upload/download/get_buckets traffic of a tahoe
     grid will cause the prefixdir contents to be mostly cached in the kernel,
     or that the number of buckets in each prefixdir will be small enough to
-    load quickly. A 1TB allmydata.com server was measured to have 2.56M
+    load quickly. A 1TB allmydata.com server was measured to have 2.56 * 10^6
     buckets, spread into the 1024 prefixdirs, with about 2500 buckets per
     prefix. On this server, each prefixdir took 130ms-200ms to list the first
     time, and 17ms to list the second time.
hunk ./src/allmydata/storage/crawler.py 68
     cpu_slice = 1.0 # use up to 1.0 seconds before yielding
     minimum_cycle_time = 300 # don't run a cycle faster than this
 
-    def __init__(self, server, statefile, allowed_cpu_percentage=None):
+    def __init__(self, backend, statefile, allowed_cpu_percentage=None):
         service.MultiService.__init__(self)
         if allowed_cpu_percentage is not None:
             self.allowed_cpu_percentage = allowed_cpu_percentage
hunk ./src/allmydata/storage/crawler.py 72
-        self.server = server
-        self.sharedir = server.sharedir
-        self.statefile = statefile
+        self.backend = backend
         self.prefixes = [si_b2a(struct.pack(">H", i << (16-10)))[:2]
                          for i in range(2**10)]
         self.prefixes.sort()
hunk ./src/allmydata/storage/crawler.py 446
 
     minimum_cycle_time = 60*60 # we don't need this more than once an hour
 
-    def __init__(self, server, statefile, num_sample_prefixes=1):
-        ShareCrawler.__init__(self, server, statefile)
+    def __init__(self, statefile, num_sample_prefixes=1):
+        ShareCrawler.__init__(self, statefile)
         self.num_sample_prefixes = num_sample_prefixes
 
     def add_initial_state(self):
hunk ./src/allmydata/storage/expirer.py 15
     removed.
 
     I collect statistics on the leases and make these available to a web
-    status page, including::
+    status page, including:
 
     Space recovered during this cycle-so-far:
      actual (only if expiration_enabled=True):
hunk ./src/allmydata/storage/expirer.py 51
     slow_start = 360 # wait 6 minutes after startup
     minimum_cycle_time = 12*60*60 # not more than twice per day
 
-    def __init__(self, server, statefile, historyfile,
+    def __init__(self, statefile, historyfile,
                  expiration_enabled, mode,
                  override_lease_duration, # used if expiration_mode=="age"
                  cutoff_date, # used if expiration_mode=="cutoff-date"
hunk ./src/allmydata/storage/expirer.py 71
         else:
             raise ValueError("GC mode '%s' must be 'age' or 'cutoff-date'" % mode)
         self.sharetypes_to_expire = sharetypes
-        ShareCrawler.__init__(self, server, statefile)
+        ShareCrawler.__init__(self, statefile)
 
     def add_initial_state(self):
         # we fill ["cycle-to-date"] here (even though they will be reset in
hunk ./src/allmydata/storage/immutable.py 44
     sharetype = "immutable"
 
     def __init__(self, filename, max_size=None, create=False):
-        """ If max_size is not None then I won't allow more than max_size to be written to me. If create=True and max_size must not be None. """
+        """ If max_size is not None then I won't allow more than
+        max_size to be written to me. If create=True then max_size
+        must not be None. """
         precondition((max_size is not None) or (not create), max_size, create)
         self.home = filename
         self._max_size = max_size
hunk ./src/allmydata/storage/immutable.py 87
 
     def read_share_data(self, offset, length):
         precondition(offset >= 0)
-        # reads beyond the end of the data are truncated. Reads that start
-        # beyond the end of the data return an empty string. I wonder why
-        # Python doesn't do the following computation for me?
+        # Reads beyond the end of the data are truncated. Reads that start
+        # beyond the end of the data return an empty string.
         seekpos = self._data_offset+offset
         fsize = os.path.getsize(self.home)
         actuallength = max(0, min(length, fsize-seekpos))
hunk ./src/allmydata/storage/immutable.py 198
             space_freed += os.stat(self.home)[stat.ST_SIZE]
             self.unlink()
         return space_freed
+class NullBucketWriter(Referenceable):
+    implements(RIBucketWriter)
 
hunk ./src/allmydata/storage/immutable.py 201
+    def remote_write(self, offset, data):
+        return
 
 class BucketWriter(Referenceable):
     implements(RIBucketWriter)
hunk ./src/allmydata/storage/server.py 7
 from twisted.application import service
 
 from zope.interface import implements
-from allmydata.interfaces import RIStorageServer, IStatsProducer
+from allmydata.interfaces import RIStorageServer, IStatsProducer, IShareStore
 from allmydata.util import fileutil, idlib, log, time_format
 import allmydata # for __full_version__
 
hunk ./src/allmydata/storage/server.py 16
 from allmydata.storage.lease import LeaseInfo
 from allmydata.storage.mutable import MutableShareFile, EmptyShare, \
      create_mutable_sharefile
-from allmydata.storage.immutable import ShareFile, BucketWriter, BucketReader
+from allmydata.storage.immutable import ShareFile, NullBucketWriter, BucketWriter, BucketReader
 from allmydata.storage.crawler import BucketCountingCrawler
 from allmydata.storage.expirer import LeaseCheckingCrawler
 
hunk ./src/allmydata/storage/server.py 20
+from zope.interface import implements
+
+# A Backend is a MultiService so that its server's crawlers (if the server has any) can
+# be started and stopped.
+class Backend(service.MultiService):
+    implements(IStatsProducer)
+    def __init__(self):
+        service.MultiService.__init__(self)
+
+    def get_bucket_shares(self):
+        """XXX"""
+        raise NotImplementedError
+
+    def get_share(self):
+        """XXX"""
+        raise NotImplementedError
+
+    def make_bucket_writer(self):
+        """XXX"""
+        raise NotImplementedError
+
+class NullBackend(Backend):
+    def __init__(self):
+        Backend.__init__(self)
+
+    def get_available_space(self):
+        return None
+
+    def get_bucket_shares(self, storage_index):
+        return set()
+
+    def get_share(self, storage_index, sharenum):
+        return None
+
+    def make_bucket_writer(self, storage_index, shnum, max_space_per_bucket, lease_info, canary):
+        return NullBucketWriter()
+
+class FSBackend(Backend):
+    def __init__(self, storedir, readonly=False, reserved_space=0):
+        Backend.__init__(self)
+
+        self._setup_storage(storedir, readonly, reserved_space)
+        self._setup_corruption_advisory()
+        self._setup_bucket_counter()
+        self._setup_lease_checkerf()
+
+    def _setup_storage(self, storedir, readonly, reserved_space):
+        self.storedir = storedir
+        self.readonly = readonly
+        self.reserved_space = int(reserved_space)
+        if self.reserved_space:
+            if self.get_available_space() is None:
+                log.msg("warning: [storage]reserved_space= is set, but this platform does not support an API to get disk statistics (statvfs(2) or GetDiskFreeSpaceEx), so this reservation cannot be honored",
+                        umid="0wZ27w", level=log.UNUSUAL)
+
+        self.sharedir = os.path.join(self.storedir, "shares")
+        fileutil.make_dirs(self.sharedir)
+        self.incomingdir = os.path.join(self.sharedir, 'incoming')
+        self._clean_incomplete()
+
+    def _clean_incomplete(self):
+        fileutil.rm_dir(self.incomingdir)
+        fileutil.make_dirs(self.incomingdir)
+
+    def _setup_corruption_advisory(self):
+        # we don't actually create the corruption-advisory dir until necessary
+        self.corruption_advisory_dir = os.path.join(self.storedir,
+                                                    "corruption-advisories")
+
+    def _setup_bucket_counter(self):
+        statefile = os.path.join(self.storedir, "bucket_counter.state")
+        self.bucket_counter = BucketCountingCrawler(statefile)
+        self.bucket_counter.setServiceParent(self)
+
+    def _setup_lease_checkerf(self):
+        statefile = os.path.join(self.storedir, "lease_checker.state")
+        historyfile = os.path.join(self.storedir, "lease_checker.history")
+        self.lease_checker = LeaseCheckingCrawler(statefile, historyfile,
+                                   expiration_enabled, expiration_mode,
+                                   expiration_override_lease_duration,
+                                   expiration_cutoff_date,
+                                   expiration_sharetypes)
+        self.lease_checker.setServiceParent(self)
+
+    def get_available_space(self):
+        if self.readonly:
+            return 0
+        return fileutil.get_available_space(self.storedir, self.reserved_space)
+
+    def get_bucket_shares(self, storage_index):
+        """Return a list of (shnum, pathname) tuples for files that hold
+        shares for this storage_index. In each tuple, 'shnum' will always be
+        the integer form of the last component of 'pathname'."""
+        storagedir = os.path.join(self.sharedir, storage_index_to_dir(storage_index))
+        try:
+            for f in os.listdir(storagedir):
+                if NUM_RE.match(f):
+                    filename = os.path.join(storagedir, f)
+                    yield (int(f), filename)
+        except OSError:
+            # Commonly caused by there being no buckets at all.
+            pass
+
 # storage/
 # storage/shares/incoming
 #   incoming/ holds temp dirs named $START/$STORAGEINDEX/$SHARENUM which will
hunk ./src/allmydata/storage/server.py 143
     name = 'storage'
     LeaseCheckerClass = LeaseCheckingCrawler
 
-    def __init__(self, storedir, nodeid, reserved_space=0,
-                 discard_storage=False, readonly_storage=False,
+    def __init__(self, nodeid, backend, reserved_space=0,
+                 readonly_storage=False,
                  stats_provider=None,
                  expiration_enabled=False,
                  expiration_mode="age",
hunk ./src/allmydata/storage/server.py 155
         assert isinstance(nodeid, str)
         assert len(nodeid) == 20
         self.my_nodeid = nodeid
-        self.storedir = storedir
-        sharedir = os.path.join(storedir, "shares")
-        fileutil.make_dirs(sharedir)
-        self.sharedir = sharedir
-        # we don't actually create the corruption-advisory dir until necessary
-        self.corruption_advisory_dir = os.path.join(storedir,
-                                                    "corruption-advisories")
-        self.reserved_space = int(reserved_space)
-        self.no_storage = discard_storage
-        self.readonly_storage = readonly_storage
         self.stats_provider = stats_provider
         if self.stats_provider:
             self.stats_provider.register_producer(self)
hunk ./src/allmydata/storage/server.py 158
-        self.incomingdir = os.path.join(sharedir, 'incoming')
-        self._clean_incomplete()
-        fileutil.make_dirs(self.incomingdir)
         self._active_writers = weakref.WeakKeyDictionary()
hunk ./src/allmydata/storage/server.py 159
+        self.backend = backend
+        self.backend.setServiceParent(self)
         log.msg("StorageServer created", facility="tahoe.storage")
 
hunk ./src/allmydata/storage/server.py 163
-        if reserved_space:
-            if self.get_available_space() is None:
-                log.msg("warning: [storage]reserved_space= is set, but this platform does not support an API to get disk statistics (statvfs(2) or GetDiskFreeSpaceEx), so this reservation cannot be honored",
-                        umin="0wZ27w", level=log.UNUSUAL)
-
         self.latencies = {"allocate": [], # immutable
                           "write": [],
                           "close": [],
hunk ./src/allmydata/storage/server.py 174
                           "renew": [],
                           "cancel": [],
                           }
-        self.add_bucket_counter()
-
-        statefile = os.path.join(self.storedir, "lease_checker.state")
-        historyfile = os.path.join(self.storedir, "lease_checker.history")
-        klass = self.LeaseCheckerClass
-        self.lease_checker = klass(self, statefile, historyfile,
-                                   expiration_enabled, expiration_mode,
-                                   expiration_override_lease_duration,
-                                   expiration_cutoff_date,
-                                   expiration_sharetypes)
-        self.lease_checker.setServiceParent(self)
 
     def __repr__(self):
         return "<StorageServer %s>" % (idlib.shortnodeid_b2a(self.my_nodeid),)
hunk ./src/allmydata/storage/server.py 178
 
-    def add_bucket_counter(self):
-        statefile = os.path.join(self.storedir, "bucket_counter.state")
-        self.bucket_counter = BucketCountingCrawler(self, statefile)
-        self.bucket_counter.setServiceParent(self)
-
     def count(self, name, delta=1):
         if self.stats_provider:
             self.stats_provider.count("storage_server." + name, delta)
hunk ./src/allmydata/storage/server.py 233
             kwargs["facility"] = "tahoe.storage"
         return log.msg(*args, **kwargs)
 
-    def _clean_incomplete(self):
-        fileutil.rm_dir(self.incomingdir)
-
     def get_stats(self):
         # remember: RIStatsProvider requires that our return dict
         # contains numeric values.
hunk ./src/allmydata/storage/server.py 269
             stats['storage_server.total_bucket_count'] = bucket_count
         return stats
 
-    def get_available_space(self):
-        """Returns available space for share storage in bytes, or None if no
-        API to get this information is available."""
-
-        if self.readonly_storage:
-            return 0
-        return fileutil.get_available_space(self.storedir, self.reserved_space)
-
     def allocated_size(self):
         space = 0
         for bw in self._active_writers:
hunk ./src/allmydata/storage/server.py 276
         return space
 
     def remote_get_version(self):
-        remaining_space = self.get_available_space()
+        remaining_space = self.backend.get_available_space()
         if remaining_space is None:
             # We're on a platform that has no API to get disk stats.
             remaining_space = 2**64
hunk ./src/allmydata/storage/server.py 301
         self.count("allocate")
         alreadygot = set()
         bucketwriters = {} # k: shnum, v: BucketWriter
-        si_dir = storage_index_to_dir(storage_index)
-        si_s = si_b2a(storage_index)
 
hunk ./src/allmydata/storage/server.py 302
+        si_s = si_b2a(storage_index)
         log.msg("storage: allocate_buckets %s" % si_s)
 
         # in this implementation, the lease information (including secrets)
hunk ./src/allmydata/storage/server.py 316
 
         max_space_per_bucket = allocated_size
 
-        remaining_space = self.get_available_space()
+        remaining_space = self.backend.get_available_space()
         limited = remaining_space is not None
         if limited:
             # this is a bit conservative, since some of this allocated_size()
hunk ./src/allmydata/storage/server.py 329
         # they asked about: this will save them a lot of work. Add or update
         # leases for all of them: if they want us to hold shares for this
         # file, they'll want us to hold leases for this file.
-        for (shnum, fn) in self._get_bucket_shares(storage_index):
+        for (shnum, fn) in self.backend.get_bucket_shares(storage_index):
             alreadygot.add(shnum)
             sf = ShareFile(fn)
             sf.add_or_renew_lease(lease_info)
hunk ./src/allmydata/storage/server.py 335
 
         for shnum in sharenums:
-            incominghome = os.path.join(self.incomingdir, si_dir, "%d" % shnum)
-            finalhome = os.path.join(self.sharedir, si_dir, "%d" % shnum)
-            if os.path.exists(finalhome):
+            share = self.backend.get_share(storage_index, shnum)
+
+            if not share:
+                if (not limited) or (remaining_space >= max_space_per_bucket):
+                    # ok! we need to create the new share file.
+                    bw = self.backend.make_bucket_writer(storage_index, shnum,
+                                      max_space_per_bucket, lease_info, canary)
+                    bucketwriters[shnum] = bw
+                    self._active_writers[bw] = 1
+                    if limited:
+                        remaining_space -= max_space_per_bucket
+                else:
+                    # bummer! not enough space to accept this bucket
+                    pass
+
+            elif share.is_complete():
                 # great! we already have it. easy.
                 pass
hunk ./src/allmydata/storage/server.py 353
-            elif os.path.exists(incominghome):
+            elif not share.is_complete():
                 # Note that we don't create BucketWriters for shnums that
                 # have a partial share (in incoming/), so if a second upload
                 # occurs while the first is still in progress, the second
hunk ./src/allmydata/storage/server.py 359
                 # uploader will use different storage servers.
                 pass
-            elif (not limited) or (remaining_space >= max_space_per_bucket):
-                # ok! we need to create the new share file.
-                bw = BucketWriter(self, incominghome, finalhome,
-                                  max_space_per_bucket, lease_info, canary)
-                if self.no_storage:
-                    bw.throw_out_all_data = True
-                bucketwriters[shnum] = bw
-                self._active_writers[bw] = 1
-                if limited:
-                    remaining_space -= max_space_per_bucket
-            else:
-                # bummer! not enough space to accept this bucket
-                pass
-
-        if bucketwriters:
-            fileutil.make_dirs(os.path.join(self.sharedir, si_dir))
 
         self.add_latency("allocate", time.time() - start)
         return alreadygot, bucketwriters
hunk ./src/allmydata/storage/server.py 437
             self.stats_provider.count('storage_server.bytes_added', consumed_size)
         del self._active_writers[bw]
 
-    def _get_bucket_shares(self, storage_index):
-        """Return a list of (shnum, pathname) tuples for files that hold
-        shares for this storage_index. In each tuple, 'shnum' will always be
-        the integer form of the last component of 'pathname'."""
-        storagedir = os.path.join(self.sharedir, storage_index_to_dir(storage_index))
-        try:
-            for f in os.listdir(storagedir):
-                if NUM_RE.match(f):
-                    filename = os.path.join(storagedir, f)
-                    yield (int(f), filename)
-        except OSError:
-            # Commonly caused by there being no buckets at all.
-            pass
 
     def remote_get_buckets(self, storage_index):
         start = time.time()
hunk ./src/allmydata/storage/server.py 444
         si_s = si_b2a(storage_index)
         log.msg("storage: get_buckets %s" % si_s)
         bucketreaders = {} # k: sharenum, v: BucketReader
-        for shnum, filename in self._get_bucket_shares(storage_index):
+        for shnum, filename in self.backend.get_bucket_shares(storage_index):
             bucketreaders[shnum] = BucketReader(self, filename,
                                                 storage_index, shnum)
         self.add_latency("get", time.time() - start)
hunk ./src/allmydata/test/test_backends.py 10
 import mock
 
 # This is the code that we're going to be testing.
-from allmydata.storage.server import StorageServer
+from allmydata.storage.server import StorageServer, FSBackend, NullBackend
 
 # The following share file contents was generated with
 # storage.immutable.ShareFile from Tahoe-LAFS v1.8.2
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())
+
+        self.failIf(mockisdir.called)
+        self.failIf(mocklistdir.called)
+        self.failIf(mockopen.called)
+        self.failIf(mockmkdir.called)
+
+        # You passed!
+
+    @mock.patch('time.time')
+    @mock.patch('os.mkdir')
     @mock.patch('__builtin__.open')
hunk ./src/allmydata/test/test_backends.py 44
-    def test_create_server(self, mockopen):
-        """ This tests whether a server instance can be constructed. """
+    @mock.patch('os.listdir')
+    @mock.patch('os.path.isdir')
+    def test_create_server_fs_backend(self, mockisdir, mocklistdir, mockopen, mockmkdir, mocktime):
+        """ This tests whether a server instance can be constructed
+        with a filesystem backend. To pass the test, it has to use the
+        filesystem in only the prescribed ways. """
 
         def call_open(fname, mode):
             if fname == 'testdir/bucket_counter.state':
hunk ./src/allmydata/test/test_backends.py 58
                 raise IOError(2, "No such file or directory: 'testdir/lease_checker.state'")
             elif fname == 'testdir/lease_checker.history':
                 return StringIO()
+            else:
+                self.fail("Server with FS backend tried to open '%s' in mode '%s'" % (fname, mode))
         mockopen.side_effect = call_open
 
         # Now begin the test.
hunk ./src/allmydata/test/test_backends.py 63
-        s = StorageServer('testdir', 'testnodeidxxxxxxxxxx')
+        s = StorageServer('testnodeidxxxxxxxxxx', backend=FSBackend('teststoredir'))
+
+        self.failIf(mockisdir.called)
+        self.failIf(mocklistdir.called)
+        self.failIf(mockopen.called)
+        self.failIf(mockmkdir.called)
+        self.failIf(mocktime.called)
 
         # You passed!
 
hunk ./src/allmydata/test/test_backends.py 73
-class TestServer(unittest.TestCase, ReallyEqualMixin):
+class TestServerNullBackend(unittest.TestCase, ReallyEqualMixin):
+    def setUp(self):
+        self.s = StorageServer('testnodeidxxxxxxxxxx', backend=NullBackend())
+
+    @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):
+        """ Write a new share. """
+
+        # Now begin the test.
+        alreadygot, bs = self.s.remote_allocate_buckets('teststorage_index', 'x'*32, 'y'*32, set((0,)), 1, mock.Mock())
+        bs[0].remote_write(0, 'a')
+        self.failIf(mockisdir.called)
+        self.failIf(mocklistdir.called)
+        self.failIf(mockopen.called)
+        self.failIf(mockmkdir.called)
+
+    @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. """
+
+        # Now begin the test.
+        bs = self.s.remote_get_buckets('teststorage_index')
+
+        self.failUnlessEqual(len(bs), 0)
+        self.failIf(mocklistdir.called)
+        self.failIf(mockopen.called)
+        self.failIf(mockgetsize.called)
+        self.failIf(mockexists.called)
+
+
+class TestServerFSBackend(unittest.TestCase, ReallyEqualMixin):
     @mock.patch('__builtin__.open')
     def setUp(self, mockopen):
         def call_open(fname, mode):
hunk ./src/allmydata/test/test_backends.py 126
                 return StringIO()
         mockopen.side_effect = call_open
 
-        self.s = StorageServer('testdir', 'testnodeidxxxxxxxxxx')
-
+        self.s = StorageServer('testnodeidxxxxxxxxxx', backend=FSBackend('teststoredir'))
 
     @mock.patch('time.time')
     @mock.patch('os.mkdir')
hunk ./src/allmydata/test/test_backends.py 134
     @mock.patch('os.listdir')
     @mock.patch('os.path.isdir')
     def test_write_share(self, mockisdir, mocklistdir, mockopen, mockmkdir, mocktime):
-        """Handle a report of corruption."""
+        """ Write a new share. """
 
         def call_listdir(dirname):
             self.failUnlessReallyEqual(dirname, 'testdir/shares/or/orsxg5dtorxxeylhmvpws3temv4a')
hunk ./src/allmydata/test/test_backends.py 173
         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)
 
hunk ./src/allmydata/test/test_backends.py 176
-
     @mock.patch('os.path.exists')
     @mock.patch('os.path.getsize')
     @mock.patch('__builtin__.open')
hunk ./src/allmydata/test/test_backends.py 218
 
         self.failUnlessEqual(len(bs), 1)
         b = bs[0]
+        # These should match by definition, the next two cases cover cases without (completely) unambiguous behaviors.
         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)
hunk ./src/allmydata/test/test_backends.py 224
         # If you start reading past the end of the file you get the empty string.
         self.failUnlessReallyEqual(b.remote_read(datalen+1, 3), '')
+
+
}

Context:

[Rename test_package_initialization.py to (much shorter) test_import.py .
Brian Warner <warner@lothar.com>**20110611190234
 Ignore-this: 3eb3dbac73600eeff5cfa6b65d65822
 
 The former name was making my 'ls' listings hard to read, by forcing them
 down to just two columns.
] 
[tests: fix tests to accomodate [20110611153758-92b7f-0ba5e4726fb6318dac28fb762a6512a003f4c430]
zooko@zooko.com**20110611163741
 Ignore-this: 64073a5f39e7937e8e5e1314c1a302d1
 Apparently none of the two authors (stercor, terrell), three reviewers (warner, davidsarah, terrell), or one committer (me) actually ran the tests. This is presumably due to #20.
 fixes #1412
] 
[wui: right-align the size column in the WUI
zooko@zooko.com**20110611153758
 Ignore-this: 492bdaf4373c96f59f90581c7daf7cd7
 Thanks to Ted "stercor" Rolle Jr. and Terrell Russell.
 fixes #1412
] 
[docs: three minor fixes
zooko@zooko.com**20110610121656
 Ignore-this: fec96579eb95aceb2ad5fc01a814c8a2
 CREDITS for arc for stats tweak
 fix link to .zip file in quickstart.rst (thanks to ChosenOne for noticing)
 English usage tweak
] 
[docs/running.rst: fix stray HTML (not .rst) link noticed by ChosenOne.
david-sarah@jacaranda.org**20110609223719
 Ignore-this: fc50ac9c94792dcac6f1067df8ac0d4a
] 
[server.py:  get_latencies now reports percentiles _only_ if there are sufficient observations for the interpretation of the percentile to be unambiguous.
wilcoxjg@gmail.com**20110527120135
 Ignore-this: 2e7029764bffc60e26f471d7c2b6611e
 interfaces.py:  modified the return type of RIStatsProvider.get_stats to allow for None as a return value
 NEWS.rst, stats.py: documentation of change to get_latencies
 stats.rst: now documents percentile modification in get_latencies
 test_storage.py:  test_latencies now expects None in output categories that contain too few samples for the associated percentile to be unambiguously reported.
 fixes #1392
] 
[docs: revert link in relnotes.txt from NEWS.rst to NEWS, since the former did not exist at revision 5000.
david-sarah@jacaranda.org**20110517011214
 Ignore-this: 6a5be6e70241e3ec0575641f64343df7
] 
[docs: convert NEWS to NEWS.rst and change all references to it.
david-sarah@jacaranda.org**20110517010255
 Ignore-this: a820b93ea10577c77e9c8206dbfe770d
] 
[docs: remove out-of-date docs/testgrid/introducer.furl and containing directory. fixes #1404
david-sarah@jacaranda.org**20110512140559
 Ignore-this: 784548fc5367fac5450df1c46890876d
] 
[scripts/common.py: don't assume that the default alias is always 'tahoe' (it is, but the API of get_alias doesn't say so). refs #1342
david-sarah@jacaranda.org**20110130164923
 Ignore-this: a271e77ce81d84bb4c43645b891d92eb
] 
[setup: don't catch all Exception from check_requirement(), but only PackagingError and ImportError
zooko@zooko.com**20110128142006
 Ignore-this: 57d4bc9298b711e4bc9dc832c75295de
 I noticed this because I had accidentally inserted a bug which caused AssertionError to be raised from check_requirement().
] 
[M-x whitespace-cleanup
zooko@zooko.com**20110510193653
 Ignore-this: dea02f831298c0f65ad096960e7df5c7
] 
[docs: fix typo in running.rst, thanks to arch_o_median
zooko@zooko.com**20110510193633
 Ignore-this: ca06de166a46abbc61140513918e79e8
] 
[relnotes.txt: don't claim to work on Cygwin (which has been untested for some time). refs #1342
david-sarah@jacaranda.org**20110204204902
 Ignore-this: 85ef118a48453d93fa4cddc32d65b25b
] 
[relnotes.txt: forseeable -> foreseeable. refs #1342
david-sarah@jacaranda.org**20110204204116
 Ignore-this: 746debc4d82f4031ebf75ab4031b3a9
] 
[replace remaining .html docs with .rst docs
zooko@zooko.com**20110510191650
 Ignore-this: d557d960a986d4ac8216d1677d236399
 Remove install.html (long since deprecated).
 Also replace some obsolete references to install.html with references to quickstart.rst.
 Fix some broken internal references within docs/historical/historical_known_issues.txt.
 Thanks to Ravi Pinjala and Patrick McDonald.
 refs #1227
] 
[docs: FTP-and-SFTP.rst: fix a minor error and update the information about which version of Twisted fixes #1297
zooko@zooko.com**20110428055232
 Ignore-this: b63cfb4ebdbe32fb3b5f885255db4d39
] 
[munin tahoe_files plugin: fix incorrect file count
francois@ctrlaltdel.ch**20110428055312
 Ignore-this: 334ba49a0bbd93b4a7b06a25697aba34
 fixes #1391
] 
[corrected "k must never be smaller than N" to "k must never be greater than N"
secorp@allmydata.org**20110425010308
 Ignore-this: 233129505d6c70860087f22541805eac
] 
[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:
f5f17113e7ee758a831726c346edff9b6ed62c2a