Subject: Re: [netatalk-admins] Re: Sherlock takes a LONG time
From: Tom Kunz (tkunz@nscs.fast.net)
Date: Tue Nov 23 1999 - 17:51:51 EST
ME wrote:
>
> The obvious tech support questions:
> ...
In a word: No. The obvious questions have an obvious answer. That's
why I'm posting here. For the not-so-obvious answers. No changes to
either clients or server. No upgrades, no changes, no nothing. I've
already done the silly tech support blah blah blah and eliminated all
that stuff. Wouldn't I be a horrificly lazy fool to not exhaust all the
"obvious tech support" possiblities before posting here? Something did
not *necessarily* change on the network to cause this problem. Latent,
hidden bugs are present in many software endeavors, thus nothing *had*
to change to cause this.
Which brings me to netatalk. The magic number of roughly 2G worth of
used disk space sounds like it *might* be some sort of arbitrary limit,
or perhaps imposed by accidental mis-prototyping of a variable. I know
others here are using much larger systems with much more space, but
perhaps that rpm contains code with a boo-boo. Also, NT servers on the
same subnet are using some apple-sharing type of product, similar in
functionality to netatalk, but with no performance problems, which says
that it is isolated to the single machine, not universally a problem
throughout the net. Like I said, file saving and copying to/from the
server are fine, also indicating some kind of problem in how Sherlock
interacts with netatalk. It happens *only* with *Sherlock*, and not
with other stuff. Things like Samba, ftpd, telnetd, remote X
connections, and so forth have no problems. It's not that Sherlock
never returns, it's that Sherlock takes a darn long time, like up to 2
hours, to search an 8G volume with 2G of files in it. It's not an issue
of SCSI vs IDE, as even an MFM or ST-506 disk won't have *THAT* much
performance difference!
Tom
-- Tom Kunz Tool Developer Software Consulting Services PGP Key http://www.users.fast.net/~tkunz/pgp.html 1452 1F99 E2BB 632E 6EAE 2DF0 EF11 4DFC DB62 7EBC 3BA0 6C40 88C0 C509 DA85 91B4 D5E9 EFD3
This archive was generated by hypermail 2b28 : Sat Dec 18 1999 - 16:17:29 EST