-
Notifications
You must be signed in to change notification settings - Fork 134
/
drain.py
320 lines (260 loc) · 12.2 KB
/
drain.py
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
"""
Description : This file implements the Drain algorithm for log parsing
Author : LogPAI team
Modified by : [email protected], [email protected]
License : MIT
"""
from typing import List, Dict
from cachetools import LRUCache, Cache
from drain3.simple_profiler import Profiler, NullProfiler
class LogCluster:
__slots__ = ["log_template_tokens", "cluster_id", "size"]
def __init__(self, log_template_tokens: list, cluster_id: int):
self.log_template_tokens = tuple(log_template_tokens)
self.cluster_id = cluster_id
self.size = 1
def get_template(self):
return ' '.join(self.log_template_tokens)
def __str__(self):
return f"ID={str(self.cluster_id).ljust(5)} : size={str(self.size).ljust(10)}: {self.get_template()}"
class Node:
__slots__ = ["key_to_child_node", "cluster_ids"]
def __init__(self):
self.key_to_child_node: Dict[str, Node] = {}
self.cluster_ids: List[int] = []
class Drain:
def __init__(self,
depth=4,
sim_th=0.4,
max_children=100,
max_clusters=None,
extra_delimiters=(),
profiler: Profiler = NullProfiler(),
param_str="<*>"):
"""
Attributes
----------
depth : depth of all leaf nodes (nodes that contain log clusters)
sim_th : similarity threshold - if percentage of similar tokens for a log message is below this
number, a new log cluster will be created.
max_children : max number of children of an internal node
max_clusters : max number of tracked clusters (unlimited by default).
When this number is reached, model starts replacing old clusters
with a new ones according to the LRU policy.
extra_delimiters: delimiters to apply when splitting log message into words (in addition to whitespace).
"""
self.depth = depth - 2 # number of prefix tokens in each tree path (exclude root and leaf node)
self.sim_th = sim_th
self.max_children = max_children
self.root_node = Node()
self.profiler = profiler
self.extra_delimiters = extra_delimiters
self.max_clusters = max_clusters
self.param_str = param_str
# key: int, value: LogCluster
self.id_to_cluster = {} if max_clusters is None else LRUCache(maxsize=max_clusters)
self.clusters_counter = 0
@property
def clusters(self):
return self.id_to_cluster.values()
@staticmethod
def has_numbers(s):
return any(char.isdigit() for char in s)
def tree_search(self, root_node: Node, tokens: list, sim_th: float, include_params: bool):
# at first level, children are grouped by token (word) count
token_count = len(tokens)
parent_node = root_node.key_to_child_node.get(str(token_count))
# no template with same token count yet
if parent_node is None:
return None
# handle case of empty log string - return the single cluster in that group
if token_count == 0:
return self.id_to_cluster.get(parent_node.cluster_ids[0])
# find the leaf node for this log - a path of nodes matching the first N tokens (N=tree depth)
current_depth = 1
for token in tokens:
# at_max_depth
if current_depth == self.depth:
break
# is_last_token
if current_depth == token_count:
break
key_to_child_node = parent_node.key_to_child_node
parent_node = key_to_child_node.get(token)
if parent_node is None:
parent_node = key_to_child_node.get(self.param_str)
if parent_node is None:
return None
current_depth += 1
# get best match among all clusters with same prefix, or None if no match is above sim_th
cluster = self.fast_match(parent_node.cluster_ids, tokens, sim_th, include_params)
return cluster
def add_seq_to_prefix_tree(self, root_node, cluster: LogCluster):
token_count_str = str(len(cluster.log_template_tokens))
if token_count_str not in root_node.key_to_child_node:
first_layer_node = Node()
root_node.key_to_child_node[token_count_str] = first_layer_node
else:
first_layer_node = root_node.key_to_child_node[token_count_str]
parent_node = first_layer_node
# handle case of empty log string
if len(cluster.log_template_tokens) == 0:
parent_node.cluster_ids = [cluster.cluster_id]
return
current_depth = 1
for token in cluster.log_template_tokens:
# if at max depth or this is last token in template - add current log cluster to the leaf node
if current_depth == self.depth or str(current_depth) == token_count_str:
# clean up stale clusters before adding a new one.
new_cluster_ids = [cluster.cluster_id]
for cluster_id in parent_node.cluster_ids:
if cluster_id in self.id_to_cluster:
new_cluster_ids.append(cluster_id)
parent_node.cluster_ids = new_cluster_ids
break
# if token not matched in this layer of existing tree.
if token not in parent_node.key_to_child_node:
if not self.has_numbers(token):
if self.param_str in parent_node.key_to_child_node:
if len(parent_node.key_to_child_node) < self.max_children:
new_node = Node()
parent_node.key_to_child_node[token] = new_node
parent_node = new_node
else:
parent_node = parent_node.key_to_child_node[self.param_str]
else:
if len(parent_node.key_to_child_node) + 1 < self.max_children:
new_node = Node()
parent_node.key_to_child_node[token] = new_node
parent_node = new_node
elif len(parent_node.key_to_child_node) + 1 == self.max_children:
new_node = Node()
parent_node.key_to_child_node[self.param_str] = new_node
parent_node = new_node
else:
parent_node = parent_node.key_to_child_node[self.param_str]
else:
if self.param_str not in parent_node.key_to_child_node:
new_node = Node()
parent_node.key_to_child_node[self.param_str] = new_node
parent_node = new_node
else:
parent_node = parent_node.key_to_child_node[self.param_str]
# if the token is matched
else:
parent_node = parent_node.key_to_child_node[token]
current_depth += 1
# seq1 is template
def get_seq_distance(self, seq1, seq2, include_params: bool):
assert len(seq1) == len(seq2)
sim_tokens = 0
param_count = 0
for token1, token2 in zip(seq1, seq2):
if token1 == self.param_str:
param_count += 1
continue
if token1 == token2:
sim_tokens += 1
if include_params:
sim_tokens += param_count
ret_val = float(sim_tokens) / len(seq1)
return ret_val, param_count
def fast_match(self, cluster_ids: list, tokens: list, sim_th: float, include_params: bool):
"""
Find the best match for a log message (represented as tokens) versus a list of clusters
:param cluster_ids: List of clusters to match against (represented by their IDs)
:param tokens: the log message, separated to tokens.
:param sim_th: minimum required similarity threshold (None will be returned in no clusters reached it)
:param include_params: consider tokens matched to wildcard parameters in similarity treshold.
:return: Best match cluster or None
"""
match_cluster = None
max_sim = -1
max_param_count = -1
max_cluster = None
for cluster_id in cluster_ids:
# Try to retrieve cluster from cache with bypassing eviction
# algorithm as we are only testing candidates for a match.
cluster = Cache.get(self.id_to_cluster, cluster_id)
if cluster is None:
continue
cur_sim, param_count = self.get_seq_distance(cluster.log_template_tokens, tokens, include_params)
if cur_sim > max_sim or (cur_sim == max_sim and param_count > max_param_count):
max_sim = cur_sim
max_param_count = param_count
max_cluster = cluster
if max_sim >= sim_th:
match_cluster = max_cluster
return match_cluster
def create_template(self, seq1, seq2):
assert len(seq1) == len(seq2)
ret_val = list(seq2)
for i, (token1, token2) in enumerate(zip(seq1, seq2)):
if token1 != token2:
ret_val[i] = self.param_str
return ret_val
def print_tree(self, file=None):
self.print_node("root", self.root_node, 0, file)
def print_node(self, token, node, depth, file):
out_str = '\t' * depth
if depth < 2:
out_str += '<' + str(token) + '>'
else:
out_str += token
print(out_str, file=file)
for token, child in node.key_to_child_node.items():
self.print_node(token, child, depth + 1, file)
def get_content_as_tokens(self, content):
content = content.strip()
for delimiter in self.extra_delimiters:
content = content.replace(delimiter, " ")
content_tokens = content.split()
return content_tokens
def add_log_message(self, content: str):
content_tokens = self.get_content_as_tokens(content)
if self.profiler:
self.profiler.start_section("tree_search")
match_cluster = self.tree_search(self.root_node, content_tokens, self.sim_th, False)
if self.profiler:
self.profiler.end_section()
# Match no existing log cluster
if match_cluster is None:
if self.profiler:
self.profiler.start_section("create_cluster")
self.clusters_counter += 1
cluster_id = self.clusters_counter
match_cluster = LogCluster(content_tokens, cluster_id)
self.id_to_cluster[cluster_id] = match_cluster
self.add_seq_to_prefix_tree(self.root_node, match_cluster)
update_type = "cluster_created"
# Add the new log message to the existing cluster
else:
if self.profiler:
self.profiler.start_section("cluster_exist")
new_template_tokens = self.create_template(content_tokens, match_cluster.log_template_tokens)
if tuple(new_template_tokens) == match_cluster.log_template_tokens:
update_type = "none"
else:
match_cluster.log_template_tokens = tuple(new_template_tokens)
update_type = "cluster_template_changed"
match_cluster.size += 1
# Touch cluster to update its state in the cache.
self.id_to_cluster.get(match_cluster.cluster_id)
if self.profiler:
self.profiler.end_section()
return match_cluster, update_type
def match(self, content: str):
"""
Match against an already existing cluster. Match shall be perfect (sim_th=1.0).
New cluster will not be created as a result of this call, nor any cluster modifications.
:param content: log message to match
:return: Matched cluster or None of no match found.
"""
content_tokens = self.get_content_as_tokens(content)
match_cluster = self.tree_search(self.root_node, content_tokens, 1.0, True)
return match_cluster
def get_total_cluster_size(self):
size = 0
for c in self.id_to_cluster.values():
size += c.size
return size