Re: [squid-users] Re: False disk full errors

On Wed, Oct 07, 1998 at 06:40:21PM +0200, Toens Bueker wrote:

> Ok. We ran into the same problem lately. Obviously there is a bug in
> solaris 2.6. Sun said you could overcome the problem by setting a bigger
> blocksize for newfs - but blocksizes bigger than 4k are not supported for
> sun4u architectures :-( You can go with only 50% disk-usage (i guess Sun
> opted for that to sell more disks) - or leave Solaris and go with Linux.

Sun told me that you could fix this by changing the numbers of bytes per
inode. The default is 2048, they suggested that this should be around 8192.

"mkfs" appears to be getting the settings very wrong, in any case. Not only
is it choosing the wrong numbers of cylinders per group, but it is also
getting maxcontig wrong ( -C 7 on the newfs line fixes this ) and it looks
like it's getting the rpm of the drives wrong, too!

I'm having this problem now with filesystems on 2Gb disks, so it looks like
something somewhere is pretty bust.

I'm suprised Sun told you that you should set a bigger blocksize than 4096
for the Sun4u architecture -- you can't use filesystems with a blocksize
_less_ than 8192 bytes on the Sun4u, as that's the page size!

I should be able to tell whether any of these tweaks will have helped within
the next 2 days, in any event, so I'll follow up to the list.

Oh, and you could, I suspect, ditch UFS and go with VxFS. It just might be a
bit pricey ...

Cheers,

Chris

-- 
Chris Tilbury, UNIX Systems Administrator, IT Services, University of Warwick
EMAIL: cudch+s@csv.warwick.ac.uk PHONE: +44 1203 523365(V)/+44 1203 523267(F)
                            URL: http://www.warwick.ac.uk/staff/Chris.Tilbury