forked from Kami/node-cassandra-client
-
Notifications
You must be signed in to change notification settings - Fork 0
/
CHANGES
144 lines (94 loc) · 5.09 KB
/
CHANGES
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
Changes with cassandra-client 0.13.0
- Fix a bug when an invalid parameter passed to PooledConnection.execute method
would cause connection pool to invalidly be exhausted instead of passing error
directly back to the user.
- Fix a bug with incorrectly waiting for a query timeout instead of imediatelly
calling 'Connection.execute' callback if an 'error' event was emitted on a
Thrift connection.
Changes with cassandra-client 0.12.0
- Remove concept of a 'taken' connection and allow the same Thrift connection to
be used to execute multiple queries at the same time.
Thrift client supports request multiplexing which means marking connection as
'taken' when a query is executing is not necessary.
- Rename 'max_query' Connection / PooledConnection option to 'query_timeout'.
Also decrease default 'query_timeout' from 100 seconds to a more sane value of
5 seconds.
- Properly propagate 'query_timeout' setting to a ConnectionInPool class inside
a PooledConnection class.
Changes with cassandra-client in 0.11.0
- Changes to support cassandra 1.1.
Mainly changes related to using binary strings instead of Buffer objects.
Regenerated Thrift client.
Still compatible with Cassandra 1.0.x.
No CQ3 yet.
Changes with cassandra-client in 0.10.0
- Pass 'metadata' object as a third and last argument to the
{Connection,PooledConnection}.execute callback.
Note: This change might break some of the existing code (e.g. if you use
connection.execute directly in async.waterfall chain) because previously
execute callback took two arguments and now it takes three.
Changes with cassandra-client in 0.9.6
- Add metadata.connectionInfo attribute to the 'rows' object passed to the
{Connection,PooledConnection}.execute callback. This object contains host and
port of the Cassandra host where the query was actually executed.
Changes with cassandra-client in 0.9.5
- fix exception thrown with undefinded query parameters [Simon Gaeremynck]
Changes with cassandra-client in 0.9.4
- support for parmeterized types (patch contributed by Allan Carroll <[email protected]>)
- log strings instead of buffers.
- Fix potential double callback on connect. [Allan Carroll]
Changes with cassandra-client in 0.9.3:
- Improve serialization and deserialization support for the following types: Number, Date, Boolean.
[Dan Foody]
Changes with cassandra-client 0.9.2:
- Fix a race condition when pooled connections are closed.
Changes with cassandra-client 0.9.1:
- Emit a 'log' event with 'error' level if an error happens when connecting to
Cassandra.
Changes with cassandra-client 0.9.0
- Change PooledConnection to retry a query if a TimedOutException is returned.
- Change Connection and PooledConnection 'log' event arguments - now it takes the
following arguments:
- level - log level
- message - string log message which is always present (previously this
argument could be either string or an object)
- obj - optional object which extra message attributes (e.g. connectionInfo,
query, etc.)
Note: This change is backward incompatible - previously 'obj' argument wasn't
passed to the 'log' event
Changes with cassandra-client 0.8.2
- id connections.
- introduce options.max_query to allow separate and configurable query timeouts.
- emit error event on thrift error.
- long query watch logs a trace when any query takes more than 10s.
- pooled connections need to quietly tear down transport connections when being reconnected.
- fix recoverable error logic in PooledConnection.execute.
- Make PooledConnection.execute() re-execution calls explicit (saw some strange inheritance problems when execute()
was overloaded with different parameters).
Changes with cassandra-client 0.8.1
- fix callback problem with CP.shutdown
Changes with cassandra-client 0.8.0
- Put drain back. (http://code.google.com/a/apache-extras.org/p/cassandra-node/issues/detail?id=32)
- Allow blob updates (http://code.google.com/a/apache-extras.org/p/cassandra-node/issues/detail?id=28)
- Remove logmagic dependency, see README.md for how to capture logging events.
(http://code.google.com/a/apache-extras.org/p/cassandra-node/issues/detail?id=16)
Changes with cassandra-client 0.7.1
- Swap out generic-pool with homegrown solution (issue 15).
- Add more cassandra types to the decoder [Christoph Tavan]
Changes with cassandra-client 0.6.2:
- Fix scope leaks.
- Replace uuid-js with node-uuid library and add a new UUID type.
[Christoph Tavan]
Changes with cassandra-client 0.6.1:
- Attach 'connectionInfo' object to the error object which is passed to the
callbacks.
Changes with cassandra-client 0.6.0:
- Set a timeout for login, learn and use steps. Defaults to 1000ms, 2000ms and
1000ms respectively.
- Set a timeout for connecting to the cassandra server. Default timeout
is 4000 ms, but a user can specify custom one by passing 'timeout'
option to the PooledConnection / Connection constructor.
- Use a different route for logging CQL queries
(node-cassandra-client.driver.cql).
- Add 'log_time' option and log query execution time if this option is true.
Timing is logged to 'node-cassandra-client.driver.timing' route.