| 1 | Revision history for AnyEvent::MP |
1 | Revision history for AnyEvent::MP |
| 2 | |
2 | |
| 3 | TODO: testsuite |
3 | TODO: testsuite |
| 4 | TODO: intro: maybe simple job pool example? |
4 | TODO: intro: maybe simple job pool example? |
| 5 | TODO: intro: mention watchdog... |
5 | TODO: intro: mention watchdog... |
| 6 | TODO: use gvpe method of making contact, if possible. |
6 | TODO: maybe disbale warnings by default? |
|
|
7 | TODO: networkdomains |
|
|
8 | TODO: group proxy nodes - store group membership for other nodes |
|
|
9 | TODO: switch to AnyEvent::Log |
| 7 | |
10 | |
|
|
11 | faq: can't se anything |
|
|
12 | faq: all is asynch |
|
|
13 | faq: how to interface to non-perl nodes? |
|
|
14 | |
|
|
15 | TODO: Anyevent::MP gets grp_reg, grp_get |
|
|
16 | TODO: AnyEvent::MP keeps list of all "global" nodes |
|
|
17 | TODO: Global loaded on demand only, adds grp database and node address database |
|
|
18 | TODO: transport connects use any global node to find addresses (maybe itself) |
|
|
19 | TODO: not really |
|
|
20 | TODO: configure norc => 1, force => 1 |
|
|
21 | |
|
|
22 | - move seed code to Kernel. |
|
|
23 | - fixed small typo in AnyEvent::MP::Transport. |
|
|
24 | |
|
|
25 | 1.30 Thu Jun 30 11:30:39 CEST 2011 |
|
|
26 | - connection errors at the right time would kill node |
|
|
27 | connections in a bad way (patch by Malcolm Studd). |
|
|
28 | |
|
|
29 | 1.29 Fri May 7 20:13:39 CEST 2010 |
|
|
30 | - codename "Sadrak". |
|
|
31 | - fix error in callback at AnyEvent/MP/Global.pm line 339 |
|
|
32 | (found by Sadrak). |
|
|
33 | - listener-less nodes had trouble sending keepalive |
|
|
34 | messages on write timeouts (found by Sadrak). |
|
|
35 | - the monitor guard could cause a memleak due |
|
|
36 | to autovivification if a mon was cleared after |
|
|
37 | the port was gone (analysed by Sadrak). |
|
|
38 | - do not overwrite the config file if we couldn't read it |
|
|
39 | for some reason (lesson demonstrated by Sadrak). |
|
|
40 | |
|
|
41 | 1.28 Thu Apr 1 21:23:54 CEST 2010 |
|
|
42 | - accepted connections didn't correctly set up the SRCNODE |
|
|
43 | in some cases, leading to intra-node messages (such as monitoring) |
|
|
44 | to fail silently (great testcase by Sadrak). |
|
|
45 | |
|
|
46 | 1.27 Sat Mar 20 21:11:16 CET 2010 |
|
|
47 | - avoid exceptions in the transport code if the connection |
|
|
48 | fails during TLS negotiation. |
|
|
49 | - delay new address broadcasts in seednodes, to reduce the |
|
|
50 | likelyhood of nodes connecting at the same time, leading to |
|
|
51 | message loss. |
|
|
52 | - improve general message passing efficiency by 8%, and special |
|
|
53 | case json framing for a 50% speed increase. |
|
|
54 | - add hints how to interface to aemp nodes from other programming |
|
|
55 | languages using the aemp protocol in AnyEvent::MP::Transport |
|
|
56 | (it can really be quite simple). |
|
|
57 | - handle unsupported remote framings more gracefully. |
|
|
58 | |
|
|
59 | 1.26 Wed Dec 30 16:48:37 CET 2009 |
|
|
60 | - INCOMPATIBLE CHANGE: the high-level node protocol changed, |
|
|
61 | all nodes need to upgrade. |
|
|
62 | - remote kil's did not work (reported by Eugene Grigoriev). |
|
|
63 | - documented monitor_timeout, connect_interval, auth_offer, |
|
|
64 | auth_accept, framing_format, autocork and nodelay options |
|
|
65 | and make them configurable via bin/aemp. |
|
|
66 | - some doc updates. |
|
|
67 | - remote kil's were not synchronous to message sends which affected |
|
|
68 | nobody, as remote kils were broken. |
|
|
69 | - remote mon/unmonitor requests were not synchronous to message |
|
|
70 | sends either sometimes, although it likely didnt matter. |
|
|
71 | |
|
|
72 | 1.24 Wed Dec 9 14:59:52 CET 2009 |
|
|
73 | - codename "the foul egg". |
|
|
74 | - increased "unmonitored local port died" message level to 2, so |
|
|
75 | it is displayed by default now. |
|
|
76 | - generate random seed at configure time, not module loading time, |
|
|
77 | so fork&configure works as expected. |
|
|
78 | - decreased "seed"-related message priority in Global from 0 (d'oh) |
|
|
79 | to 9, as they are not signifcant. |
|
|
80 | - new experimental module AnyEvent::MP::DataConn to establish |
|
|
81 | additional data connections between nodes. |
|
|
82 | - bin/aemp now loads AnyEvent::Watchdog::Util before trying |
|
|
83 | to call restart. |
|
|
84 | - bin/aemp setcert didn't properly handle the filename argument. |
|
|
85 | - removed empty and unused timeout= parameter from protocol greeting. |
|
|
86 | |
|
|
87 | 1.23 Fri Nov 6 18:46:26 CET 2009 |
|
|
88 | - listener-less nodes were misinformed by their masters about |
|
|
89 | the location of new nodes, and therefore didn't connect |
|
|
90 | properly in all cases. |
|
|
91 | - messages send during node-up processing could sometimes get lost, |
|
|
92 | which would hamper global's ability to mesh the network. |
| 8 | - fixed AnyEvent::MP::Global::grp_reg to return a guard also |
93 | - fixed AnyEvent::MP::Global::grp_reg to return a guard also |
|
|
94 | in scalar context. |
|
|
95 | - fixed AnyEvent::MP::Kernel::mon_nodes to return a guard also |
| 9 | in scalar context. |
96 | in scalar context. |
| 10 | - try to improve error reporting when automatically loading |
97 | - try to improve error reporting when automatically loading |
| 11 | a function - being unable to load a module will now |
98 | a function - being unable to load a module will now |
| 12 | stop the process and report the error. |
99 | stop the process and report the error. |
| 13 | |
100 | |
