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
|
Changelog for s6.
In 2.7.2.0
----------
- s6-ipcserver-socketbinder now takes a -B option to make the socket blocking.
- s6-svscan now takes a "-d notif" option to notify readiness. This is
*shallow readiness*, i.e. readiness will be reported when s6-svscan is ready
to receive s6-svscanctl commands; it does *not* mean all the services started
at s6-svscan launch will be ready.
- New optional file in a service directory: down-signal. It allows you to
customize the signal sent by s6-svc -d (and now s6-svc -r too).
- New s6-svc option: -r. Sends the signal defined in the down-signal file
(still SIGTERM by default).
- Adaptation to skalibs-2.7.0.0.
In 2.7.1.1
----------
- Bugfixes.
In 2.7.1.0
----------
- Bugfixes.
- s6-svwait et al. now exit 102 instead of hanging if s6-supervise dies.
- New commands: s6-svdt, s6-svdt-clear, s6-permafailon
- s6-tai64nlocal can now print GMT times with the -g option.
In 2.7.0.0
----------
- Bugfixes.
- s6-ftrigrd and s6lockd internally use the textclient API/ABI
instead of skaclient. This is cleaner, faster code.
- New "-a perms" option to s6-ipcserver and s6-ipcserver-socketbinder.
In 2.6.2.0
----------
- Bugfixes.
- New functions: s6_fdholder_start() and s6_fdholder_end().
In 2.6.1.1
----------
- Bugfixes.
In 2.6.1.0
----------
- Bugfixes.
- Addition of the s6_svc_lock_take() and s6_svc_lock_release()
functions, to fix a race condition in s6-rc.
- New program: s6-notifyoncheck, to implement polling for
readiness and reroute the result into the s6 readiness notification
mechanism.
In 2.6.0.0
----------
- Bugfixes. One of the bugfixes required a visible ABI change,
which explains the major version bump.
- Documentation fixes.
In 2.5.1.0
----------
- Bugfixes.
- s6-ftrig-listen1 now prints to stdout the last event (the last
byte) it received and made it match the regexp it was given.
- s6-supervise now recognizes the "timeout-kill" file in a service
directory. If this file exists and contains a nonzero integer t,
the behaviour of a "s6-svc -d" command is changed from
"send a SIGTERM and a SIGCONT" to "send a SIGTERM and a SIGCONT,
then if the service is still not dead after t milliseconds, send
a SIGKILL".
In 2.5.0.0
----------
- Added this NEWS file. :)
- Major types overhaul to make them more POSIXly correct:
compatibility with skalibs-2.5.0.0.
- Services can now report permanent failure by having a ./finish
script exit 125.
- s6-svwait, s6-svlisten and s6-svlisten1 now exit the number of
services that reported permanent failure. They exit 99 on timeout.
- When a service has a ./nosetsid file that contains "setpgrp", it
is run in a new process group, but not in a new session. If the
./nosetsid file contains anything else, the service is run in the
same session and process group as the supervisor, and if there is
no such file, it's run in a new session (same behaviour as before).
- s6-log now writes a warning to stderr when it closes the stdout
stream.
- s6-log's individual file size limit bumped to 256 MB.
- Various bugfixes. (thanks jjk!)
|
