aboutsummaryrefslogtreecommitdiffstats
path: root/show-blocks
blob: 49e1770e5bd0bf4f0233fda1e59f3a91eba28e31 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
#!/usr/bin/env python
#
# Copyright (C) 2007 Oracle.  All rights reserved.
#
# This program is free software; you can redistribute it and/or
# modify it under the terms of the GNU General Public
# License v2 as published by the Free Software Foundation.
# 
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
# General Public License for more details.
# 
# You should have received a copy of the GNU General Public
# License along with this program; if not, write to the
# Free Software Foundation, Inc., 59 Temple Place - Suite 330,
# Boston, MA 021110-1307, USA.
#
import sys, os, signal, time, commands, tempfile, random

# numpy seems to override random() with something else.  Instantiate our
# own here
randgen = random.Random()
randgen.seed(50)

from optparse import OptionParser
from matplotlib import rcParams
from matplotlib.font_manager import fontManager, FontProperties
import numpy

rcParams['numerix'] = 'numpy'
rcParams['backend'] = 'Agg'
rcParams['interactive'] = 'False'
from pylab import *

class AnnoteFinder:
  """
  callback for matplotlib to display an annotation when points are clicked on.  The
  point which is closest to the click and within xtol and ytol is identified.
    
  Register this function like this:
    
  scatter(xdata, ydata)
  af = AnnoteFinder(xdata, ydata, annotes)
  connect('button_press_event', af)
  """

  def __init__(self, axis=None):
    if axis is None:
      self.axis = gca()
    else:
      self.axis= axis
    self.drawnAnnotations = {}
    self.links = []
    
  def clear(self):
    for k in self.drawnAnnotations.keys():
        self.drawnAnnotations[k].set_visible(False)

  def __call__(self, event):
    if event.inaxes:
      if event.button != 1:
        self.clear()
        draw()
        return
      clickX = event.xdata
      clickY = event.ydata
      if (self.axis is None) or (self.axis==event.inaxes):
        self.drawAnnote(event.inaxes, clickX, clickY)
    
  def drawAnnote(self, axis, x, y):
    """
    Draw the annotation on the plot
    """
    if self.drawnAnnotations.has_key((x,y)):
      markers = self.drawnAnnotations[(x,y)]
      markers.set_visible(not markers.get_visible())
      draw()
    else:
      t = axis.text(x,y, "(%3.2f, %3.2f)"%(x,y), bbox=dict(facecolor='red',
                    alpha=0.8))
      self.drawnAnnotations[(x,y)] = t
      draw()

def loaddata(fh,delimiter=None, converters=None):

    #14413824 8192 extent back ref root 5 gen 10 owner 282 num_refs 1
    def iter(fh, delimiter, converters):
        global total_data
        global total_metadata
        for i,line in enumerate(fh):
            line = line.split(' ')
            start = float(line[0])
            len = float(line[1])
            owner = float(line[10])
            root = float(line[6])
            if owner <= 255:
                total_metadata += int(len)
            else:
                total_data += int(len)
            if start < zoommin or (zoommax != 0 and start > zoommax):
                continue
            yield start
            yield len
            yield owner
            yield root
    X = numpy.fromiter(iter(fh, delimiter, converters), dtype=float)
    return X

def run_debug_tree(device):
    p = os.popen('btrfs-debug-tree -e ' + device)
    data = loaddata(p)
    return data

def shapeit(X):
    lines = len(X) / 4
    X.shape = (lines, 4)

def line_picker(line, mouseevent):
    if mouseevent.xdata is None: return False, dict()
    print "%d %d\n", mouseevent.xdata, mouseevent.ydata
    return False, dict()

def xycalc(byte):
    byte = byte / bytes_per_cell
    yval = floor(byte / num_cells)
    xval = byte % num_cells
    return (xval, yval + 1)

# record the color used for each root the first time we find it
root_colors = {}
# there are lots of good colormaps to choose from
# http://www.scipy.org/Cookbook/Matplotlib/Show_colormaps
#
meta_cmap = get_cmap("gist_ncar")
data_done = False

def plotone(a, xvals, yvals, owner, root, lines, labels):
    global data_done
    add_label = False

    if owner:
        if options.meta_only:
            return
        color = "blue"
        label = "Data"
        if not data_done:
            add_label = True
            data_done = True
    else:
        if options.data_only:
            return
        if root not in root_colors:
            color = meta_cmap(randgen.random())
            label = "Meta %d" % int(root)
            root_colors[root] = (color, label)
            add_label = True
        else:
            color, label = root_colors[root]

    plotlines = a.plot(xvals, yvals, 's', color=color, mfc=color, mec=color,
           markersize=.23, label=label)
    if add_label:
        lines += plotlines
        labels.append(label)
        print "add label %s" % label

def parse_zoom():
    def parse_num(s):
        mult = 1
        c = s.lower()[-1]
        if c == 't':
            mult = 1024 * 1024 * 1024 * 1024
        elif c == 'g':
            mult = 1024 * 1024 * 1024
        elif c == 'm':
            mult = 1024 * 1024
        elif c == 'k':
            mult = 1024
        else:
            c = None
        if c:
            num = int(s[:-1]) * mult
        else:
            num = int(s)
        return num
        
    if not options.zoom:
        return (0, 0)

    vals = options.zoom.split(':')
    if len(vals) != 2:
        sys.stderr.write("warning: unable to parse zoom %s\n" % options.zoom)
        return (0, 0)
    zoommin = parse_num(vals[0])
    zoommax = parse_num(vals[1])
    return (zoommin, zoommax)

usage = "usage: %prog [options]"
parser = OptionParser(usage=usage)
parser.add_option("-d", "--device", help="Btrfs device", default="")
parser.add_option("-i", "--input-file", help="debug-tree data", default="")
parser.add_option("-o", "--output", help="Output file", default="blocks.png")
parser.add_option("-z", "--zoom", help="Zoom", default=None)
parser.add_option("", "--data-only", help="Only print data blocks",
                  default=False, action="store_true")
parser.add_option("", "--meta-only", help="Only print metadata blocks",
                  default=False, action="store_true")

(options,args) = parser.parse_args()

if not options.device and not options.input_file:
    parser.print_help()
    sys.exit(1)

zoommin, zoommax = parse_zoom()
total_data = 0
total_metadata = 0

if options.device:
    data = run_debug_tree(options.device)
elif options.input_file:
    data = loaddata(file(options.input_file))
shapeit(data)

# try to drop out the least common data points by creating
# a histogram of the sectors seen.
sectors = data[:,0]
sizes = data[:,1]
datalen = len(data)
sectormax = numpy.max(sectors)
sectormin = 0
num_cells = 800
total_cells = num_cells * num_cells
byte_range = sectormax - sectormin
bytes_per_cell = byte_range / total_cells

f = figure(figsize=(8,6))

# Throughput goes at the bottom
a = subplot(1, 1, 1)
subplots_adjust(right=0.7)
datai = 0
xvals = []
yvals = []
last_owner = 0
last_root = 0
lines = []
labels = []
while datai < datalen:
    row = data[datai]
    datai += 1
    byte = row[0]
    size = row[1]
    owner = row[2]
    root = row[3]

    if owner <= 255:
        owner = 0
    else:
        owner = 1

    if len(xvals) and (owner != last_owner or last_root != root):
        plotone(a, xvals, yvals, last_owner, last_root, lines, labels)
        xvals = []
        yvals = []
    cell = 0
    while cell < size:
        xy = xycalc(byte)
        byte += bytes_per_cell
        cell += bytes_per_cell
        if xy:
            xvals.append(xy[0])
            yvals.append(xy[1])
    last_owner = owner
    last_root = root

if xvals:
    plotone(a, xvals, yvals, last_owner, last_root, lines, labels)

# make sure the final second goes on the x axes
ticks = []
a.set_xticks(ticks)
ticks = a.get_yticks()

first_tick = ticks[1] * bytes_per_cell * num_cells
if first_tick > 1024 * 1024 * 1024 * 1024:
    scale = 1024 * 1024 * 1024 * 1024;
    scalestr = "TB"
elif first_tick > 1024 * 1024 * 1024:
    scale = 1024 * 1024 * 1024;
    scalestr = "GB"
elif first_tick > 1024 * 1024:
    scale = 1024 * 1024;
    scalestr = "MB"
elif first_tick > 1024:
    scale = 1024;
    scalestr = "KB"
else:
    scalestr = "Bytes"
    scale = 1

ylabels = [ str(int((x * bytes_per_cell * num_cells) / scale)) for x in ticks ]
a.set_yticklabels(ylabels)
a.set_ylabel('Disk offset (%s)' % scalestr)
a.set_xlim(0, num_cells)
a.set_title('Blocks')

a.legend(lines, labels, loc=(1.05, 0.8), shadow=True, pad=0.1, numpoints=1,
              handletextsep = 0.005,
              labelsep = 0.01,
              markerscale=10,
              prop=FontProperties(size='x-small') )

if total_data == 0:
    percent_meta = 100
else:
    percent_meta = (float(total_metadata) / float(total_data)) * 100

print "Total metadata bytes %d data %d ratio %.3f" % (total_metadata,
                                                    total_data, percent_meta)
print "saving graph to %s" % options.output
savefig(options.output, orientation='landscape')
show()