Re: Re: Client timeout - renegotiation failed - connection dead
scunacc <[email protected]> Sat, 04 Mar 2006 07:33:18 -0500
| Newsgroups | gmane.linux.enbd.general |
|---|---|
| Message-ID | <1141475598.4432.233.camel@scuna-gate> |
Dear Peter, > > > Since a couple of people are having time problems, I took the opportunity > > > to replace the calls to gettimeofday in time.c with calls to the > > > "secret" uptime function in libproc, > Well, I've had to make it more sophisticated, since every so often the > procps library gets > > bad data in /proc/uptime > > so I still have to fall back to system time, dynamically, and thus > might be vulnerable to system time changes. That's not good is it. Wonder why procps does that... Hmmm. > > Did you make the kinds of changes we had with the code as well? Or will > > I have to do those again? Just curious. > > I think I did. I just flushed what was on my machine to the server. > It's a develpment snapshot! That's great - I'll grab and try. - Did you change the makefile for make config all so that it does a userspace build with a different KERNELDIR than the module build? Or does that still have to be done in two separate stages still as we did "by hand" before? > > I hope to go back to enbd too, but interestingly, nbd worked, and has > > been working without failure overnight. > I don't think it checks times at all (it's vulnerable to getting stuck > forever). Aaaah. OK. Neat thing is, (I think - can't see a downside here), I can use enbd as a drop in replacement when I've got the new version built and tested as everything else is the same. Kind regards Derek.