forked from cockroachdb/cockroach
/
doc.go
149 lines (112 loc) · 6.41 KB
/
doc.go
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
// Copyright 2016 The Cockroach Authors.
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
// implied. See the License for the specific language governing
// permissions and limitations under the License.
//
// Author: Matt Tracy (matt.r.tracy@gmail.com)
/*
Package ts provides a basic time series datastore on top of the underlying
CockroachDB key/value datastore. It is used to serve basic metrics generated by
CockroachDB.
Storing time series data is a unique challenge for databases. Time series data
is typically generated at an extremely high volume, and is queried by providing
a range of time of arbitrary size, which can lead to an enormous amount of data
being scanned for a query. Many specialized time series databases already exist
to meet these challenges; those solutions are built on top of specialized
storage engines which are often unsuitable for general data storage needs, but
currently superior to CockroachDB for the purpose of time series data.
However, it is a broad goal of CockroachDB to provide a good experience for
developers, and out-of-the-box recording of internal metrics is helpful for
small and prototype deployments.
Organization Structure
Time series data is organized on disk according to two basic, sortable properties:
+ Time series name (i.e "sql.operations.selects")
+ Timestamp
This is optimized for querying data for a single series over multiple
timestamps: data for the same series at different timestamps is stored
contiguously.
Downsampling
The amount of data produced by time series sampling can be considerable; storing
every incoming data point with perfect fidelity can command a tremendous amount
of computing and storage resources.
However, in many use cases perfect fidelity is not necessary; the exact time a
sample was taken is unimportant, with the overall trend of the data over time
being far more important to analysis than the individual samples.
With this in mind, CockroachDB downsamples data before storing it; the original
timestamp for each data point in a series is not recorded. CockroachDB instead
divides time into contiguous slots of uniform length (currently 10 seconds); if
multiple data points for a series fall in the same slot, only the most recent
sample is kept.
Slab Storage
In order to use key space efficiently, we pack data for multiple contiguous
samples into "slab" values, with data for each slab stored in a CockroachDB key.
This is done by again dividing time into contiguous slots, but with a longer
duration; this is known as the "slab duration". For example, CockroachDB
downsamples its internal data at a resolution of 10 seconds, but stores it with
a "slab duration" of 1 hour, meaning that all samples that fall in the same hour
are stored at the same key. This strategy helps reduce the number of keys
scanned during a query.
Source Keys
Another common use case of time series queries is the aggregation of multiple series;
for example, you may want to query the same metric (e.g. "queries per second") across multiple machines
on a cluster, and aggregate the result.
Specialized Time-series databases can often aggregate across arbitrary series;
however, CockroachDB is specialized for aggregation of the same series across
different machines or disks.
This is done by creating a "source key", typically a node or store ID, which is
an optional identifier that is separate from the series name itself. The source
key is are appended to the key as a suffix, after the series name and timestamp;
this means that data that is from the same series and time period, but from
different nodes, will be stored contiguously in the key space. Data from all
sources in a series can thus be queried in a single scan.
Unused Feature: Multiple resolutions
CockroachDB time series database has rudimentary support for a planned feature:
recording the same series at multiple sample durations, commonly known as a
"rollup".
For example, a single series may be recorded with a sample size of 10 seconds,
but also record the same data with a sample size of 1 hour. The 1 hour data will
have much less information, but can be queried much faster; this is very useful
when querying a series over a very long period of time (e.g. an entire month or
year).
A specific sample duration in CockroachDB is known as a Resolution. CockroachDB
supports a fixed set of Resolutions; each Resolution has a fixed sample duration
and a slab duration. For example, the resolution "Resolution10s" has a sample
duration of 10 seconds and a slab duration of 1 hour.
This feature was planned and slightly informs our key structure (resolution
information is encoded in every time series key); however, all time series in
CockroachDB are currently recorded at a downsample duration of 10 seconds, and a
slab duration of 1 hour.
Example
A hypothetical example from CockroachDB: we want to record the available
capacity of all stores in the cluster.
The series name is: cockroach.capacity.available
Data points for this series are automatically collected from all stores. When data points are
written, they are recorded with a source key of: [store id]
There are 3 stores which contain data: 1, 2 and 3. These are arbitrary and may
change over time.
Data is recorded for January 1st, 2016 between 10:05 pm and 11:05 pm. The data
is recorded at a 10 second resolution.
The data is recorded into keys structurally similar to the following:
tsd.cockroach.capacity.available.10s.403234.1
tsd.cockroach.capacity.available.10s.403234.2
tsd.cockroach.capacity.available.10s.403234.3
tsd.cockroach.capacity.available.10s.403235.1
tsd.cockroach.capacity.available.10s.403235.2
tsd.cockroach.capacity.available.10s.403235.3
Data for each source is stored in two keys: one for the 10 pm hour, and one
for the 11pm hour. Each key contains the tsd prefix, the series name, the
resolution (10s), a timestamp representing the hour, and finally the series key. The
keys will appear in the data store in the order shown above.
(Note that the keys will NOT be exactly as pictured above; they will be encoded
in a way that is more efficient, but is not readily human readable.)
*/
package ts