Debian Codegraph @

CHANGES

raw
1 1.13.002 2016-07-11
2 * pod links to gearmand repaired
3
4 1.13.001 2016-07-11
5 * Feature Request: #115368 for Gearman-Server: Allow bind to specific IP address
6 * Fix bug #115458 Distributions contain no tests.
7 Actually only use ok and version tests
8 * Fix bug #89033 typo fixes
9 * Fix bug #107045 [PATCH] fix pod whatis
10 * Fix bug #115350 Uses old port by default (7003) should use 4730
11 * pod usage
12
13 1.12 2014-12-14
14
15 * Add HACKING file
16
17 * Fix bug #70728 memory leak when clients disconnect (sleeper list isn't pruned).
18 Fixes CPAN RT 70728 (Marsh Yamazaki)
19
20 1.11 2010-01-17
21
22 * Don't get stuck in the wakeup loop. Dummy.
23
24 1.10 2009-10-04
25
26 * Read client input more aggressively. Speed improvement.
27
28 * Add text commands 'jobs' and 'clients' which give information allowing you to trace
29 jobs from clients to workers while they are running.
30
31 * Flush buffers to client on EOF (assume half-closed). This makes things like netcat
32 work better as a way to speak text protocol to the gearmand.
33
34 * Add command-line option to adjust a delay before more workers are woken up.
35 This acts as an anti-starvation mechanism in case of lower wake up counts.
36 -Default is .1 seconds, formerly this option was not needed because all workers
37 were woken up at the time of job submission.
38
39 * Add command-line option to adjust number of workers to wake up per job injected.
40 -Default is 3, formerly was -1 (wake up all as fast as possible)
41
42 * Add command-line option to change the number of sockets accepted at once per
43 listener socket.
44 -Default is now 10, formerly used to be 1.
45
46 * Add exceptions passing support to gearman server classes, using new options
47 support.
48
49 * Add options support to server clients, so they can subscribe to newer protocol
50 features.
51
52 * Change listening socket to be a real Danga::Socket subclass, this allows
53 pausing for a period of time when we run into accept errors. This will
54 fix the problem of gearmand spinning 100% cpu in those cases.
55
56 * Make gearmand a little more vocal about socket accept errors.
57
58 * add fast read concept to server reading from client codepath. This drastically
59 improves performance of jobs over 1k in size.
60
61 * fix in-process client and start_worker method calls to use non-blocking
62 sockets like they should.
63
64 1.09 2007-05-09
65
66 * make start_worker (for making worker child processes), return
67 the pid in scalar context, or ($pid, $conn) in list context
68
69 * make the text command "status" show functions which have
70 registered workers, even if nothing is pending for that func.
71
72 1.08 2007-05-01
73
74 * let start_worker take a coderef for the exec, not just an $exe name
75
76 1.07 2007-05-01
77
78 * fix up bug in Gearman::Server: we set OtherFds, instead of adding with
79 Danga::Socket->AddOtherFds, blowing away other watched fds.
80
81 * Add Gearman::Server->to_inprocess_server that returns a socket connecting to
82 the server object, we can then use this socket with a Gearman::Client::Async
83 object to run clients and servers in the same thread.
84
85 * Add ability for workers to be launched as sub processes of the
86 gearmand, using a duped socketpair for communication.
87
88 1.06 2007-04-25
89
90 * split up the monolithic gearmand script (which had classes, but
91 all internally), into separate files per class, and also make
92 a new Gearman::Server class, so Gearman servers can be embedded
93 into other processes. (running Gearman router inside Perlbal,
94 MogileFS, DJabberd, etc... then having child processes be workers)
95
96 1.05
97
98 * add --pidfile=/path/file.pid option and document the command
99 line options (Ask Bjoern Hansen)
100
101 1.04 2007-02-20
102
103 * Add maximum jobqueue size feature, with a test.
104
105 1.03 2006-09-25
106
107 * "version" text command
108
109 * weakref client connections in listeners list so they go away.
110 still wondering why the jobs holding them lasted so long.
111 or was it a loop? time will tell. go gladiator.
112
113 1.02 2006-09-20
114
115 * Devel::Gladiator support for tracking elusive memory leak
116
117 * support for "shutdown" and "shutdown graceful" commands,
118 the latter of which immediate stops listens, and exits
119 when convenient
120
121 1.01 2006-07-01
122
123 * Artur Bergman backed out some changes seconds before I did
124 the 1.00 release. this puts them back in.
125
126 1.00 2006-07-01
127
128 * the get-it-on-CPAN-finally release. but forgot this changes
129 entry.
130
131 0.30 2006-06-27
132
133 * fix memory leak with sleepers list. change it from list to hash
134 so there can't be dups, as there were before, when a worker
135 could do multiple functions.