Made a small improvement to the client cache to that makes it slightly

MRUish.

In some experiments, this led to significant reductions in cache miss
rates.

src/ZEO/cache.py

@@ -152,7 +152,7 @@ class ClientCache(object):
     # default of 20MB.  The default here is misleading, though, since
     # ClientStorage is the only user of ClientCache, and it always passes an
     # explicit size of its own choosing.
-    def __init__(self, path=None, size=200*1024**2):
+    def __init__(self, path=None, size=200*1024**2, rearrange=.8):
 
         # - `path`:  filepath for the cache file, or None (in which case
         #   a temp file will be created)
@@ -163,6 +163,11 @@ class ClientCache(object):
         size = max(size, ZEC_HEADER_SIZE)
         self.maxsize = size
 
+        # rearrange: if we read a current record and it's more than
+        # rearrange*size from the end, then copy it forward to keep it
+        # from being evicted.
+        self.rearrange = rearrange * size
+
         # The number of records in the cache.
         self._len = 0
 
@@ -497,6 +502,7 @@ class ClientCache(object):
         size, saved_oid, tid, end_tid, lver, ldata = unpack(
             ">I8s8s8sHI", read(34))
         assert saved_oid == oid, (ofs, self.f.tell(), oid, saved_oid)
+        assert end_tid == z64, (ofs, self.f.tell(), oid, tid, end_tid)
         assert lver == 0, "Versions aren't supported"
 
         data = read(ldata)
@@ -508,6 +514,25 @@ class ClientCache(object):
 
         self._n_accesses += 1
         self._trace(0x22, oid, tid, end_tid, ldata)
+
+        ofsofs = self.currentofs - ofs
+        if ofsofs < 0:
+            ofsofs += self.maxsize
+
+        if (ofsofs > self.rearrange and
+            self.maxsize > 10*len(data) and
+            size > 4):
+            # The record is far back and might get evicted, but it's
+            # valuable, so move it forward.
+
+            # Remove fromn old loc:
+            del self.current[oid]
+            self.f.seek(ofs)
+            self.f.write('f'+pack(">I", size))
+
+            # Write to new location:
+            self._store(oid, tid, None, data, size)
+
         return data, tid
 
     ##
@@ -599,6 +624,16 @@ class ClientCache(object):
         self._n_added_bytes += size
         self._len += 1
 
+        self._store(oid, start_tid, end_tid, data, size)
+
+        if end_tid:
+            self._trace(0x54, oid, start_tid, end_tid, dlen=len(data))
+        else:
+            self._trace(0x52, oid, start_tid, dlen=len(data))
+
+    def _store(self, oid, start_tid, end_tid, data, size):
+        # Low-level store used by store and load
+
         # In the next line, we ask for an extra to make sure we always
         # have a free block after the new alocated block.  This free
         # block acts as a ring pointer, so that on restart, we start
@@ -618,6 +653,7 @@ class ClientCache(object):
             extra = 'f' + pack(">I", excess)
 
         ofs = self.currentofs
+        seek = self.f.seek
         seek(ofs)
         write = self.f.write
 
@@ -639,13 +675,12 @@ class ClientCache(object):
 
         if end_tid:
             self._set_noncurrent(oid, start_tid, ofs)
-            self._trace(0x54, oid, start_tid, end_tid, dlen=len(data))
         else:
             self.current[oid] = ofs
-            self._trace(0x52, oid, start_tid, dlen=len(data))
 
         self.currentofs += size
 
+
     ##
     # If `tid` is None,
     # forget all knowledge of `oid`.  (`tid` can be None only for
@@ -660,7 +695,6 @@ class ClientCache(object):
     # - oid object id
     # - tid the id of the transaction that wrote a new revision of oid,
     #        or None to forget all cached info about oid.
-
     @locked
     def invalidate(self, oid, tid):
         ofs = self.current.get(oid)

src/ZEO/scripts/cache_simul.py

@@ -19,6 +19,7 @@ Usage: simul.py [-s size] tracefile
 Options:
 -s size: cache size in MB (default 20 MB)
 -i: summarizing interval in minutes (default 15; max 60)
+-r: rearrange factor
 
 Note:
 
@@ -52,18 +53,22 @@ def main(args=None):
     # Parse options.
     MB = 1<<20
     cachelimit = 20*MB
+    rearrange = 0.8
     simclass = CircularCacheSimulation
     interval_step = 15
     try:
-        opts, args = getopt.getopt(args, "s:i:")
+        opts, args = getopt.getopt(args, "s:i:r:")
     except getopt.error, msg:
         usage(msg)
         return 2
+
     for o, a in opts:
         if o == '-s':
             cachelimit = int(float(a)*MB)
         elif o == '-i':
             interval_step = int(a)
+        elif o == '-r':
+            rearrange = float(a)
         else:
             assert False, (o, a)
 
@@ -103,8 +108,8 @@ def main(args=None):
             return 1
 
     # Create simulation object.
-    sim = simclass(cachelimit)
-    interval_sim = simclass(cachelimit)
+    sim = simclass(cachelimit, rearrange)
+    interval_sim = simclass(cachelimit, rearrange)
 
     # Print output header.
     sim.printheader()
@@ -141,6 +146,8 @@ def main(args=None):
             if last_interval is not None:
                 interval_sim.report()
                 interval_sim.restart()
+                if not interval_sim.warm:
+                    sim.restart()
             last_interval = this_interval
         sim.event(ts, dlen, version, code, oid, start_tid, end_tid)
         interval_sim.event(ts, dlen, version, code, oid, start_tid, end_tid)
@@ -160,10 +167,12 @@ class Simulation(object):
     finish() method also calls report().
     """
 
-    def __init__(self, cachelimit):
+    def __init__(self, cachelimit, rearrange):
         self.cachelimit = cachelimit
+        self.rearrange = rearrange
         # Initialize global statistics.
         self.epoch = None
+        self.warm = False
         self.total_loads = 0
         self.total_hits = 0       # subclass must increment
         self.total_invals = 0     # subclass must increment
@@ -284,18 +293,17 @@ class Simulation(object):
 
 # For use in CircularCacheSimulation.
 class CircularCacheEntry(object):
-    __slots__ = (# object key:  an (oid, start_tid) pair, where
-                 # start_tid is the tid of the transaction that created
-                 # this revision of oid
+    __slots__ = (
+        # object key: an (oid, start_tid) pair, where start_tid is the
+        # tid of the transaction that created this revision of oid
         'key',
 
-                 # tid of transaction that created the next revision;
-                 # z64 iff this is the current revision
+        # tid of transaction that created the next revision; z64 iff
+        # this is the current revision
         'end_tid',
 
-                 # Offset from start of file to the object's data
-                 # record; this includes all overhead bytes (status
-                 # byte, size bytes, etc).
+        # Offset from start of file to the object's data record; this
+        # includes all overhead bytes (status byte, size bytes, etc).
         'offset',
         )
 
@@ -316,10 +324,12 @@ class CircularCacheSimulation(Simulation):
 
     extras = "evicts", "inuse"
 
-    def __init__(self, cachelimit):
+    evicts = 0
+
+    def __init__(self, cachelimit, rearrange):
         from ZEO import cache
 
-        Simulation.__init__(self, cachelimit)
+        Simulation.__init__(self, cachelimit, rearrange)
         self.total_evicts = 0  # number of cache evictions
 
         # Current offset in file.
@@ -348,6 +358,8 @@ class CircularCacheSimulation(Simulation):
 
     def restart(self):
         Simulation.restart(self)
+        if self.evicts:
+            self.warm = True
         self.evicts = 0
         self.evicted_hit = self.evicted_miss = 0
 
@@ -358,6 +370,20 @@ class CircularCacheSimulation(Simulation):
             if oid in self.current: # else it's a cache miss
                 self.hits += 1
                 self.total_hits += 1
+
+                tid = self.current[oid]
+                entry = self.key2entry[(oid, tid)]
+                offset_offset = self.offset - entry.offset
+                if offset_offset < 0:
+                    offset_offset += self.cachelimit
+                    assert offset_offset >= 0
+
+                if offset_offset > self.rearrange * self.cachelimit:
+                    # we haven't accessed it in a while.  Move it forward
+                    size = self.filemap[entry.offset][0]
+                    self._remove(*entry.key)
+                    self.add(oid, size, tid)
+
             elif oid in self.evicted:
                 size, e = self.evicted[oid]
                 self.write(oid, size, e.key[1], z64, 1)