Date: Thu, 10 May 2007 16:51:54 +0200 From: Ivan Voras <ivoras@fer.hr> To: freebsd-current@freebsd.org Subject: Re: Experiences with 7.0-CURRENT and vmware. Message-ID: <f1vbil$juk$1@sea.gmane.org> In-Reply-To: <20070510125445.GA5460@hub.freebsd.org> References: <20070510111326.GA94093@hub.freebsd.org> <20070510132153.A91312@fledge.watson.org> <20070510125445.GA5460@hub.freebsd.org>
next in thread | previous in thread | raw e-mail | index | archive | help
This is an OpenPGP/MIME signed message (RFC 2440 and 3156) --------------enig80F24AAB267D3E2D9338AA81 Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: quoted-printable >>> I should add that FreeBSD 6, with the same setting, is no better and = that=20 >>> I need to run ntpdate every 5-10 minutes via crontab in order to keep= good=20 >>> time (timekeeping is *really* bad.) In one instance, i was watching = >>> "zpool iostat 1" and it appeared like the rows were muching up at a r= ate=20 >>> of 2 a second for a minute or so. How do I disable TSC timekeeping? = >>> (NetBSD has this disabled by default in their kernels.) Or is there = >>> somethign else I must do? >> kern.timecounter.hardware: ACPI-fast >> kern.timecounter.choice: TSC(800) ACPI-fast(1000) i8254(0) dummy(-1000= 000) I don't see the beginings of this disucssion, but, if you use TSC=20 timekeeping in guest, and the host is multicore (may also be valid for=20 multi-socket...) running Windows, the Windows scheduler will throw the=20 vmware process around on the CPUs, giving weird measurements in the=20 guest machine. The solution is to bind the vmware-vmx process in the=20 Windows task manager to one CPU only. Using this, I get what appears to=20 be "reasonable" timekeeping (didn't really fiddle with it more, but I=20 run ntpd). I also run on a very low kern.hz (50), this may or may not make a big=20 difference. >>> Second, networking. Prior to FreeBSD-7, the driver to use inside vmwa= re=20 >>> workstation was lnc. It has worked and contiues to work great. No=20 >>> problemo. FreeBSD-7 uses the "em" driver. To put it simply, it sucks= in=20 >>> comparison. When things really get bad I start seeing "em0: watchdog= =20 >>> timeout" messages on the console. I looked and I don't see a lnc dri= ver=20 >>> anywhere. Is there another alternative (le?) driver that I can use i= n=20 >>> place of em, if so, how? >> Has VMware changed what network hardware they emulate, and/or does VMw= are=20 >> offer options about what virtual hardware to expose? Recent VMWare versions can emulate either AMD Lance (or its variation,=20 it appears, called vix) or Intel "E1000" card. See, for example, my=20 writing on this: http://ivoras.net/freebsd/vmware.html. In retrospect, I think I can't really say there's a difference in=20 performance between le and em drivers in VMWare - other slowdowns=20 dominate the results. --------------enig80F24AAB267D3E2D9338AA81 Content-Type: application/pgp-signature; name="signature.asc" Content-Description: OpenPGP digital signature Content-Disposition: attachment; filename="signature.asc" -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.5 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iD8DBQFGQzGKldnAQVacBcgRAgK8AKCXtGEfQgnhcctKB4OVbJEp1LoLNwCcCVhu eesy4HIYy4zb4Gzq0hC5JjY= =U0BM -----END PGP SIGNATURE----- --------------enig80F24AAB267D3E2D9338AA81--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?f1vbil$juk$1>