hdd timeout? was share disappeares smbd/server.c:exit_server(408)

[Bernhard Ertl]

Hi,

I got a new idea. I have heared sth. about segate scsi hdds having problems with the write cache which may lead
to delays about 3 minutes (I only heared it from sinix-z).

Theory: with increasing netload the possibility of a writing operation to run into cache troubles increases.
Three minutes pause, then it works again....
(also might be totally bullshit)

Could anyone tell me what error 408 exactly means?

Does anyon know anything how to disable this write cache?

has anyone similar ( or other timeout) problems who is using segate / Ibm drives

Be

> ok, please investigate more, using nmblookup, smbclient etc, increasing
> log files etc.
>
> we have someone else who is using RAID and they also have problems.
>
> On Thu, 13 Apr 2000, Bernhard Ertl wrote:
>
> > Hi, I just recompiled it, but without any improvement, it was the gcc 2.95.2-19.
> > I also recompiled the Kernel, it is version 2.2.14. (but the suse kernel sources).
> > I also made some statistics about the occurrance of the problem and found out
> > till end of february there were agerage 25 occurrances a day (between 0 and 50)
> > but then there was a big jump to 150 to 400 /day (without haveing done anything, of course....
> >
> > Be
> >
> >
> > > On Wed, 12 Apr 2000, Bernhard Ertl wrote:
> > >
> > > > > AH!  someone else reported this happening.  GIVE INFO! details.
> > > >
> > > > > operating system arch compiler etc.
> > > >
> > > > The operating system is linux, SuSE 6.4 distribution, kernel 2.2.14 (but it also appeared with the old
> > > > version 6.2 and Kernel 2.2.?) I used the sources from the samba site, dated on November ~11th (directly
> > > > after the information about the new release came across this list.) As I think it was not available as
> > > > bindist, i suppose I have compiled it myself with gcc and glibc2. (I have to do some research about the
> > > > version, because we just have run a server update).
> > > >
> > > > The system runs on a dual PIII system with a mylex raid controller and two ethernet interfaces, serving
> > > > around 50 clients as login- wins- and fileserver.
> > > >
> > > > The problem occurrs if
> > > > * there are find processes on the hdd (understandeable)
> > > > * someone publishes a big Frontpage web located on the server (also understandeable)
> > >
> > > bernhard, thank you very much, because someone else reported a similar
> > > problem, you see.
> >
>
> <a href=" mailto:lkcl at samba.org" > Luke Kenneth Casson Leighton    </a>
> <a href=" http://cb1.com/~lkcl"  > Samba and Network Development   </a>
> <a href=" http://samba.org"      > Samba Web site                  </a>
> <a href=" http://mcp.com"        > Macmillan Technical Publishing  </a>
>
> ISBN1578701503 DCE/RPC over SMB: Samba and Windows NT Domain Internals