1 |
>>> Well there's 2 immediate differences between our setups, and I'm |
2 |
>>> disinclined to think that architecture is an issue cause I've run |
3 |
>>> amanda |
4 |
>>> on ppc as well (2.4.3 I think)., and never raced xinetd. I'm on x86 |
5 |
>>> now, |
6 |
>>> using xinetd-2.3.13 like you, but not using the ebuild. Not to be |
7 |
>>> presumptuous, but have you tried doing a manual installation? It was |
8 |
>>> a |
9 |
>>> major pain the first few times, but I can show my configs if you |
10 |
>>> want to |
11 |
>>> try that route. I have no idea whether that will fix your issue but |
12 |
>>> it |
13 |
>>> might be worth a go. Been running amanda 2.4.3 for a long while, |
14 |
>>> 2.4.4 |
15 |
>>> recently. |
16 |
>> |
17 |
>> I get exactly the same results with a manual build as I do with the |
18 |
>> ebuild. Just gave it a try again to be sure, I still get the same |
19 |
>> race condition. I also tried building amanda manually, didn't help |
20 |
>> the situation either. I've run amanda and xinetd together before on |
21 |
>> ppc but not under gentoo... I have them running together happily on |
22 |
>> an x86 box sitting right next to the 'problem' server too. I'm |
23 |
>> starting to wonder if there's a problem with some dependancy of |
24 |
>> xinetd now. <sigh> |
25 |
>> |
26 |
>> Bryn |
27 |
> |
28 |
> Just tried recompiling all dependencies of both xinetd and amanda, no |
29 |
> change. xinetd still races after the second connection. I'm pretty |
30 |
> much at a loss as to what to try next. :( |
31 |
> |
32 |
> Bryn |
33 |
|
34 |
Just tried building the old-style inetd (netkit). It seems to work |
35 |
perfectly well with Amanda. I guess I'll stick with it for the time |
36 |
being. This at least confirms that amanda itself is working fine on my |
37 |
box. Given that the only service I'm dealing with is amanda and I |
38 |
don't really need any of the features xinetd offers I guess it'll just |
39 |
have to be good enough! |
40 |
|
41 |
Bryn |
42 |
|
43 |
|
44 |
-- |
45 |
gentoo-ppc-user@g.o mailing list |