[hipl-dev] Re: [Branch ~hipl-core/hipl/trunk] Rev 5952: Require network and local filesystem to be initialized in init scripts.
- From: David Martin <david.martin.mailbox@xxxxxxxxxxxxxx>
- To: hipl-dev@xxxxxxxxxxxxx
- Date: Wed, 8 Jun 2011 15:55:21 +0200
Hi,
>> You make a good point there. I'm not sure if the settings were really
>> thought out when they were set like that.
>> - we do need the local filesystem, that's for sure, right? So that
>> should be in it.
>
> Yes because hipd tries to load the configuration files.
>
>> - we do only work on our local filesystem so $remote_fs may be scrapped?
>
> Yes. While it's possible, it's very unlikely that the configuration files
> are kept on a remote file system.
>
>> - we do use syslog for logging so this should be kept in
>
> Strictly speaking, syslog should be excluded because it's possible to do
> remote syslog. A potential problem is that I haven't tested what happens if
> you start hipd while syslog is not running yet. According to
> lib/core/debug.c:hip_vlog() always reopens the syslog channel and
> HIP_DEBUG/ERROR/etc do not have any return values, so I think this should
> work but please try this out and confirm this.
Hm ok, I've put this on my todo list for the time being. Maybe Fahad
knows more about this as he set it up if I remember correctly. I'll
get back to him on it.
>> - for the reasons you stated we may not want to wait for an active
>> network connection, so this can be removed as well
>
> This is the Right Way. However, I am not sure how well hipd survives without
> the loopback interface (used for loading configuration via hipconf trickery)
> so I recommend testing this.
I have committed the changes in trunk revision 5954 and will keep a
close look on it.
Thanks a bunch for your input!
David
Other related posts:
