| … | |
… | |
| 16 | - listener-less nodes had trouble sending keepalive |
16 | - listener-less nodes had trouble sending keepalive |
| 17 | messages on write timeouts (found by Sadrak). |
17 | messages on write timeouts (found by Sadrak). |
| 18 | - the monitor guard could cause a memleak due |
18 | - the monitor guard could cause a memleak due |
| 19 | to autovivification if a mon was cleared after |
19 | to autovivification if a mon was cleared after |
| 20 | the port was gone (analysed by Sadrak). |
20 | the port was gone (analysed by Sadrak). |
|
|
21 | - do not overwrite the config file if we couldn't read it |
|
|
22 | for some reason (lesson demonstrated by Sadrak). |
| 21 | |
23 | |
| 22 | 1.28 Thu Apr 1 21:23:54 CEST 2010 |
24 | 1.28 Thu Apr 1 21:23:54 CEST 2010 |
| 23 | - accepted connections didn't correctly set up the SRCNODE |
25 | - accepted connections didn't correctly set up the SRCNODE |
| 24 | in some cases, leading to intra-node messages (such as monitoring) |
26 | in some cases, leading to intra-node messages (such as monitoring) |
| 25 | to fail silently (great testcase by Sadrak). |
27 | to fail silently (great testcase by Sadrak). |
