[khiltd]

I’m not 100% clear on what your current configuration is, but from what I was able to gather it sounds like you’re using Apache to cover up a symptom rather than address its cause. Rather than redirecting anything I think you should probably focus on configuring your internal nameserver to provide the same information as the rest of the net–maybe delegate a zone to the dynamic host and/or use a VPN for your remote employees. I do something similar only I use my external web server’s copy of BIND to handle the DDNS updates for me (as opposed to DynDNS) and it works pretty well.

[khiltd]

That was extremely difficult to read and I’m still not sure what it is you’re asking for. You’re obviously aware of the existence of other operating systems so I have a hard time believing that anyone could possibly name a practical alternative you’ve never heard of before. What is it you’re expecting people to say in response to this?

[khiltd]

chown and chmod both do recursion with the -R option. Or is the problem that the action doesn’t fire for anything deeper?

[khiltd]

I’ve honestly never tried it through Server Admin, but fancy DNS configurations almost always involve working with the zone files directly. No GUI can present every possible configuration directive that BIND has to offer in a comprehensible format–there’s just too many of them.

[khiltd]

I think you’ve misread the documentation:

[code]-N, –timestamping don’t re-retrieve files unless newer than local.[/code]

[khiltd]

Folder actions or launchd watchpath? It’ll be slow, but free.

[khiltd]

What if you lose the -N option on wget?

[khiltd]

What’s in the script and why does it need root privileges to copy a file?

[khiltd]

DNS is always worth looking into if you want to find things by name rather than number since that’s what makes it work. The .local names don’t function outside the subnet or if the router blocks such traffic. If you don’t have anything mapping these names to addresses and if you don’t have any clients that are told where to find these maps then I’m afraid your traffic’s going nowhere.

[khiltd]

Is this server running BIND and do the client machines know about it? If so you probably have it misconfigured for what you’re trying to accomplish and the contents of a relevant zone file would be helpful.

[khiltd]

If it’s just an open file server then a clean install probably wouldn’t be all that disruptive.

[khiltd]

I would recommend either starting over with a clean install or doing an intermediate bump to 10.4 (applying all available updates) before going to 10.5. I have had nothing but problems with the two machines I’ve taken from 10.3 straight to 10.5. Others may have had better luck than I.

[khiltd]

Disk drivers don’t [i]usually[/i] log errors to disk, since:

A. that’s kind of like calling somebody to tell them their phone is broken, and

B. few logging facilities are available at interrupt time ([url=http://developer.apple.com/documentation/DeviceDrivers/Conceptual/WritingDeviceDriver/DebuggingDrivers/chapter_8_section_5.html#//apple_ref/doc/uid/TP30000701-BAJHHHDA]click for details[/url])

The drive firmware also doesn’t necessarily ever tell anyone when it decides to remap a bad sector.

If you’re particularly adventurous you can give [url=http://lists.apple.com/archives/Darwin-drivers/2007/May/msg00052.html]this[/url] a shot, but using one of the usual drive and filesystem testing utilities would probably be your best bet.

[khiltd]

If nothing changed it wouldn’t be a new version. The fact that 10.4 clients seem to get along better with a 10.4 server shouldn’t really present that big of a mystery. While the Leopard Finder does seem slower to log in to anything initially, I have not had any of the 10+ minute hangs which always resulted when a mounted share disappeared from the network to date.

[khiltd]

[QUOTE][u]Quote by: macshome[/u][p]1. PlistBuddy is a default install on 10.5 now.[/p][/QUOTE]

Ah, hiding in /usr/libexec/ I see.

[Author]

Posts