Re: Linux on BHyVe in 10.0-RELEASE
Seems like it is a processor motherboard combo thing (see other thread) On Sun, Jan 26, 2014 at 9:26 AM, Markiyan Kushnir markiyan.kush...@gmail.com wrote: yes, once I posted it I realized that these sysctls are not relevant at this stage. There was a segfault, was that bhyvectl? May be it makes sense for someone (a bhyve dev) to inspect it? -- Markiyan. 2014-01-26 Aryeh Friedman aryeh.fried...@gmail.com: On Sun, Jan 26, 2014 at 8:46 AM, Markiyan Kushnir markiyan.kush...@gmail.com wrote: hmm... 15:45:test-bhyve# md5 ubuntu-12.04.3-server-amd64.iso MD5 (ubuntu-12.04.3-server-amd64.iso) = 2cbe868812a871242cdcdd8f2fd6feb9 # md5 ubuntu/ubuntu.iso MD5 (ubuntu/ubuntu.iso) = 2cbe868812a871242cdcdd8f2fd6feb9 doing the sysctl does no good -- Aryeh M. Friedman, Lead Developer, http://www.PetiteCloud.org -- Aryeh M. Friedman, Lead Developer, http://www.PetiteCloud.org ___ freebsd-virtualization@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-virtualization To unsubscribe, send any mail to freebsd-virtualization-unsubscr...@freebsd.org
BHyVe - ESXi comparison
Hello everybody. We did a very rough comparison betweend BHyVe and VMWare ESXi. Maybe you want to give it a read and let me know if I did write a bunch of sh!t :-) http://andrea.brancatelli.it/2014/01/28/freebsd-10-0-bhyve-vmware-esxi-5-5-comparison/ I must say I'm very pleased with BHyVe performances! Very good work!! Thanks for your time. -- *Andrea BrancatelliSchema 31 S.r.l. - Socio UnicoResponsabile ITROMA - FIRENZE - PALERMO ITALYTel: +39. 06.98.358.472* *Cell: +39 331.2488468Fax: +39. 055.71.880.466Società del Gruppo SC31 ITALIA* ___ freebsd-virtualization@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-virtualization To unsubscribe, send any mail to freebsd-virtualization-unsubscr...@freebsd.org
Re: Linux on BHyVe in 10.0-RELEASE
2014-01-28 Aryeh Friedman aryeh.fried...@gmail.com: Seems like it is a processor motherboard combo thing (see other thread) ah, ok. good to know :) -- Markiyan. On Sun, Jan 26, 2014 at 9:26 AM, Markiyan Kushnir markiyan.kush...@gmail.com wrote: yes, once I posted it I realized that these sysctls are not relevant at this stage. There was a segfault, was that bhyvectl? May be it makes sense for someone (a bhyve dev) to inspect it? -- Markiyan. 2014-01-26 Aryeh Friedman aryeh.fried...@gmail.com: On Sun, Jan 26, 2014 at 8:46 AM, Markiyan Kushnir markiyan.kush...@gmail.com wrote: hmm... 15:45:test-bhyve# md5 ubuntu-12.04.3-server-amd64.iso MD5 (ubuntu-12.04.3-server-amd64.iso) = 2cbe868812a871242cdcdd8f2fd6feb9 # md5 ubuntu/ubuntu.iso MD5 (ubuntu/ubuntu.iso) = 2cbe868812a871242cdcdd8f2fd6feb9 doing the sysctl does no good -- Aryeh M. Friedman, Lead Developer, http://www.PetiteCloud.org -- Aryeh M. Friedman, Lead Developer, http://www.PetiteCloud.org ___ freebsd-virtualization@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-virtualization To unsubscribe, send any mail to freebsd-virtualization-unsubscr...@freebsd.org
Re: BHyVe - ESXi comparison
W dniu 2014-01-28 12:18, Andrea Brancatelli pisze: We did a very rough comparison betweend BHyVe and VMWare ESXi. Maybe you want to give it a read and let me know if I did write a bunch of sh!t :-) http://andrea.brancatelli.it/2014/01/28/freebsd-10-0-bhyve-vmware-esxi-5-5-comparison/ I must say I'm very pleased with BHyVe performances! Very good work!! Thanks for your time. You've wrote: BHyVe took 332 seconds VMWare took 313 seconds The difference is about 106.7%. I think that the correct conclusion should state: if we assume VMware's time as a reference than BHyVe was 6.7% slower (not 106.7%). And it should be 6,41% really :) -- best regards, Lukasz Wasikowski ___ freebsd-virtualization@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-virtualization To unsubscribe, send any mail to freebsd-virtualization-unsubscr...@freebsd.org
Re: BHyVe - ESXi comparison
And lets not forget the head start vmware has (bhyve is what? 1-2 years old?) and the size of it. Less then 1mb in code. On Tue, Jan 28, 2014 at 12:27 PM, Łukasz Wąsikowski luk...@wasikowski.netwrote: W dniu 2014-01-28 12:18, Andrea Brancatelli pisze: We did a very rough comparison betweend BHyVe and VMWare ESXi. Maybe you want to give it a read and let me know if I did write a bunch of sh!t :-) http://andrea.brancatelli.it/2014/01/28/freebsd-10-0-bhyve-vmware-esxi-5-5-comparison/ I must say I'm very pleased with BHyVe performances! Very good work!! Thanks for your time. You've wrote: BHyVe took 332 seconds VMWare took 313 seconds The difference is about 106.7%. I think that the correct conclusion should state: if we assume VMware's time as a reference than BHyVe was 6.7% slower (not 106.7%). And it should be 6,41% really :) -- best regards, Lukasz Wasikowski ___ freebsd-virtualization@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-virtualization To unsubscribe, send any mail to freebsd-virtualization-unsubscr...@freebsd.org ___ freebsd-virtualization@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-virtualization To unsubscribe, send any mail to freebsd-virtualization-unsubscr...@freebsd.org
Re: BHyVe - ESXi comparison
Yes, I tried to emphasized that both at the beginning and at the end of the post... On Tue, Jan 28, 2014 at 12:39 PM, Matthias Gamsjager mgamsja...@gmail.comwrote: And lets not forget the head start vmware has (bhyve is what? 1-2 years old?) and the size of it. Less then 1mb in code. On Tue, Jan 28, 2014 at 12:27 PM, Łukasz Wąsikowski luk...@wasikowski.net wrote: W dniu 2014-01-28 12:18, Andrea Brancatelli pisze: We did a very rough comparison betweend BHyVe and VMWare ESXi. Maybe you want to give it a read and let me know if I did write a bunch of sh!t :-) http://andrea.brancatelli.it/2014/01/28/freebsd-10-0-bhyve-vmware-esxi-5-5-comparison/ I must say I'm very pleased with BHyVe performances! Very good work!! Thanks for your time. You've wrote: BHyVe took 332 seconds VMWare took 313 seconds The difference is about 106.7%. I think that the correct conclusion should state: if we assume VMware's time as a reference than BHyVe was 6.7% slower (not 106.7%). And it should be 6,41% really :) -- best regards, Lukasz Wasikowski ___ freebsd-virtualization@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-virtualization To unsubscribe, send any mail to freebsd-virtualization-unsubscr...@freebsd.org -- *Andrea BrancatelliSchema 31 S.r.l. - Socio UnicoResponsabile ITROMA - FIRENZE - PALERMO ITALYTel: +39. 06.98.358.472* *Cell: +39 331.2488468Fax: +39. 055.71.880.466Società del Gruppo SC31 ITALIA* ___ freebsd-virtualization@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-virtualization To unsubscribe, send any mail to freebsd-virtualization-unsubscr...@freebsd.org
Re: BHyVe - ESXi comparison
W dniu 2014-01-28 12:52, Andrea Brancatelli pisze: OK, I changed that, thanks for your feedback, my assumption was something like with bhyve it took the 106% of the time it took with VMWare, but probably the approach of bhyve being 6% slower is clearer. I think so too. Thank you for testing! :) -- best regards, Lukasz Wasikowski ___ freebsd-virtualization@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-virtualization To unsubscribe, send any mail to freebsd-virtualization-unsubscr...@freebsd.org
Re: BHyVe - ESXi comparison
Fixed, thanks. On Tue, Jan 28, 2014 at 1:16 PM, Mark Martinec mark.martinec+free...@ijs.si wrote: http://andrea.brancatelli.it/2014/01/28/freebsd-10-0-bhyve- vmware-esxi-5-5-comparison/ the seconds you see is a medium of all the values from the different machines medium??? A median or an average? If you have any question on the datas Plural of datum is data, not datas. Mark ___ freebsd-virtualization@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-virtualization To unsubscribe, send any mail to freebsd-virtualization- unsubscr...@freebsd.org -- *Andrea BrancatelliSchema 31 S.r.l. - Socio UnicoResponsabile ITROMA - FIRENZE - PALERMO ITALYTel: +39. 06.98.358.472* *Cell: +39 331.2488468Fax: +39. 055.71.880.466Società del Gruppo SC31 ITALIA* ___ freebsd-virtualization@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-virtualization To unsubscribe, send any mail to freebsd-virtualization-unsubscr...@freebsd.org
Re: BHyVe - ESXi comparison
I'd have to find a different workload (compiling a port under linux makes no sense), but that something I was already thinking about. Anybody has any idea about that? It must be something that get's done the same way (so for example if we are compiling it has to be gcc vs. gcc, but gcc is not the same on the two platforms)... I don't know...? Any benchmarking suite? But I don't want to benchmark the OS that is in the middle... On Tue, Jan 28, 2014 at 3:57 PM, Lars Engels lars.eng...@0x20.net wrote: Am 2014-01-28 13:21, schrieb Andrea Brancatelli: Fixed, thanks. Could you also compare two instances of Linux inside bhyve and VMware? ___ freebsd-virtualization@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-virtualization To unsubscribe, send any mail to freebsd-virtualization- unsubscr...@freebsd.org -- *Andrea BrancatelliSchema 31 S.r.l. - Socio UnicoResponsabile ITROMA - FIRENZE - PALERMO ITALYTel: +39. 06.98.358.472* *Cell: +39 331.2488468Fax: +39. 055.71.880.466Società del Gruppo SC31 ITALIA* ___ freebsd-virtualization@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-virtualization To unsubscribe, send any mail to freebsd-virtualization-unsubscr...@freebsd.org
Re: BHyVe - ESXi comparison
Am 2014-01-28 13:21, schrieb Andrea Brancatelli: Fixed, thanks. Could you also compare two instances of Linux inside bhyve and VMware? ___ freebsd-virtualization@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-virtualization To unsubscribe, send any mail to freebsd-virtualization-unsubscr...@freebsd.org
Re: BHyVe - ESXi comparison
Oh sorry you mean linux vs. linux! Then yes, I can do that! On Tue, Jan 28, 2014 at 4:02 PM, Andrea Brancatelli abrancate...@schema31.it wrote: I'd have to find a different workload (compiling a port under linux makes no sense), but that something I was already thinking about. Anybody has any idea about that? It must be something that get's done the same way (so for example if we are compiling it has to be gcc vs. gcc, but gcc is not the same on the two platforms)... I don't know...? Any benchmarking suite? But I don't want to benchmark the OS that is in the middle... On Tue, Jan 28, 2014 at 3:57 PM, Lars Engels lars.eng...@0x20.net wrote: Am 2014-01-28 13:21, schrieb Andrea Brancatelli: Fixed, thanks. Could you also compare two instances of Linux inside bhyve and VMware? ___ freebsd-virtualization@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-virtualization To unsubscribe, send any mail to freebsd-virtualization- unsubscr...@freebsd.org -- *Andrea BrancatelliSchema 31 S.r.l. - Socio Unico Responsabile IT ROMA - FIRENZE - PALERMO ITALY Tel: +39. 06.98.358.472* *Cell: +39 331.2488468 Fax: +39. 055.71.880.466Società del Gruppo SC31 ITALIA * -- *Andrea BrancatelliSchema 31 S.r.l. - Socio UnicoResponsabile ITROMA - FIRENZE - PALERMO ITALYTel: +39. 06.98.358.472* *Cell: +39 331.2488468Fax: +39. 055.71.880.466Società del Gruppo SC31 ITALIA* ___ freebsd-virtualization@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-virtualization To unsubscribe, send any mail to freebsd-virtualization-unsubscr...@freebsd.org
Re: BHyVe - ESXi comparison
Hi Andrea, On Tue, Jan 28, 2014 at 3:18 AM, Andrea Brancatelli abrancate...@schema31.it wrote: Hello everybody. We did a very rough comparison betweend BHyVe and VMWare ESXi. Maybe you want to give it a read and let me know if I did write a bunch of sh!t :-) Looks good to me :-) http://andrea.brancatelli.it/2014/01/28/freebsd-10-0-bhyve-vmware-esxi-5-5-comparison/ I must say I'm very pleased with BHyVe performances! Very good work!! Thanks for your time. Could you see if you can reproduce the hang with 20VMs with the host running FreeBSD current? Another thing you could do is compile the host with KTR and then when you see the hangexecute the following command on the host: sudo ktrdump -crto /tmp/ktrdump.out kernel options to enable KTR: options KTR options KTR_ENTRIES=(4*1024*1024) options KTR_MASK=(KTR_GEN) best Neel -- *Andrea BrancatelliSchema 31 S.r.l. - Socio UnicoResponsabile ITROMA - FIRENZE - PALERMO ITALYTel: +39. 06.98.358.472* *Cell: +39 331.2488468Fax: +39. 055.71.880.466Società del Gruppo SC31 ITALIA* ___ freebsd-virtualization@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-virtualization To unsubscribe, send any mail to freebsd-virtualization-unsubscr...@freebsd.org ___ freebsd-virtualization@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-virtualization To unsubscribe, send any mail to freebsd-virtualization-unsubscr...@freebsd.org
Re: BHyVe - ESXi comparison
Hello Peter, unfortunately we've been a bit sloppy in tracking the time output because initially it was just an internal test, thus we don't have the details. We're setting up a new round of tests we'll run tomorrow and we'll track user/system/real in a more precise way; I will also publish a graph with the three stacked piles. Hyperthreading should hopefully be enable on the host, frankly I didn't check it out, I will tomorrow. KVM and QEMU are a bit out of our scope, so we didn't have plans for that. If I can fine some spare time we'll try. On Tue, Jan 28, 2014 at 5:10 PM, Peter Grehan gre...@freebsd.org wrote: Hi Andrea, We did a very rough comparison betweend BHyVe and VMWare ESXi. Maybe you want to give it a read and let me know if I did write a bunch of sh!t :-) Looks good to me :) Thanks for running the tests. Would you be able to list the command options you used with bhyve when running these tests ? What I couldn’t really understand (but that’s something not related to bhyve or VMWare) is how a multiprocessor machine is slower than a singleprocessor machine in doing the compilation… any idea? Is hyper-threading enabled on your system ? If not, then with a host only having 2 CPUs and a 2 vCPU guest, there isn't as much opportunity to overlap host i/o threads with vCPU threads. It would be interesting to see your time results when running bhyve to show %user/%system etc - that may give an indication of how much time is spent on 'overhead' CPU usage as opposed to pure vCPU usage. 20 VM – 2 CPUs – 2GB RAM Interesting result to say the least :) I'll try and repro this and see if it's something simple. At first guess I'd say it's the classic 'lock-holder-preemption' issue that the ESXi scheduler has a lot of smarts to avoid. Another interesting test would be Qemu/KVM VMs on Linux to see if it has the same issue. later, Peter. -- *Andrea BrancatelliSchema 31 S.r.l. - Socio UnicoResponsabile ITROMA - FIRENZE - PALERMO ITALYTel: +39. 06.98.358.472* *Cell: +39 331.2488468Fax: +39. 055.71.880.466Società del Gruppo SC31 ITALIA* ___ freebsd-virtualization@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-virtualization To unsubscribe, send any mail to freebsd-virtualization-unsubscr...@freebsd.org
Re: BHyVe - ESXi comparison
Unfortunately these are pre-production environments thus installing something fancy wasn't in our scope. If I can allocate some time and some hardware I'll try to. On Tue, Jan 28, 2014 at 5:49 PM, Neel Natu neeln...@gmail.com wrote: Hi Andrea, On Tue, Jan 28, 2014 at 3:18 AM, Andrea Brancatelli abrancate...@schema31.it wrote: Hello everybody. We did a very rough comparison betweend BHyVe and VMWare ESXi. Maybe you want to give it a read and let me know if I did write a bunch of sh!t :-) Looks good to me :-) http://andrea.brancatelli.it/2014/01/28/freebsd-10-0-bhyve-vmware-esxi-5-5-comparison/ I must say I'm very pleased with BHyVe performances! Very good work!! Thanks for your time. Could you see if you can reproduce the hang with 20VMs with the host running FreeBSD current? Another thing you could do is compile the host with KTR and then when you see the hangexecute the following command on the host: sudo ktrdump -crto /tmp/ktrdump.out kernel options to enable KTR: options KTR options KTR_ENTRIES=(4*1024*1024) options KTR_MASK=(KTR_GEN) best Neel -- *Andrea BrancatelliSchema 31 S.r.l. - Socio UnicoResponsabile ITROMA - FIRENZE - PALERMO ITALYTel: +39. 06.98.358.472* *Cell: +39 331.2488468Fax: +39. 055.71.880.466Società del Gruppo SC31 ITALIA* ___ freebsd-virtualization@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-virtualization To unsubscribe, send any mail to freebsd-virtualization-unsubscr...@freebsd.org -- *Andrea BrancatelliSchema 31 S.r.l. - Socio UnicoResponsabile ITROMA - FIRENZE - PALERMO ITALYTel: +39. 06.98.358.472* *Cell: +39 331.2488468Fax: +39. 055.71.880.466Società del Gruppo SC31 ITALIA* ___ freebsd-virtualization@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-virtualization To unsubscribe, send any mail to freebsd-virtualization-unsubscr...@freebsd.org
Re: BHyVe - ESXi comparison
OK, tomorrow I'll check. Today we tried the standard compile approach, compiling PERL to have something that would work at least a few minutes. To give you a fast anticipation, debian on bhyve took 2 minutes 23, while debian on vmware took 2 minutes and 7 seconds. Will update my post tomorrow morning with more details. On Tue, Jan 28, 2014 at 5:39 PM, Michael Berman michael.ber...@tidalscale.com wrote: parsec and stresslinux may be of interest. On 1/28/14, 7:02 AM, Andrea Brancatelli abrancate...@schema31.it wrote: I'd have to find a different workload (compiling a port under linux makes no sense), but that something I was already thinking about. Anybody has any idea about that? It must be something that get's done the same way (so for example if we are compiling it has to be gcc vs. gcc, but gcc is not the same on the two platforms)... I don't know...? Any benchmarking suite? But I don't want to benchmark the OS that is in the middle... On Tue, Jan 28, 2014 at 3:57 PM, Lars Engels lars.eng...@0x20.net wrote: Am 2014-01-28 13:21, schrieb Andrea Brancatelli: Fixed, thanks. Could you also compare two instances of Linux inside bhyve and VMware? ___ freebsd-virtualization@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-virtualization To unsubscribe, send any mail to freebsd-virtualization- unsubscr...@freebsd.org -- *Andrea BrancatelliSchema 31 S.r.l. - Socio UnicoResponsabile ITROMA - FIRENZE - PALERMO ITALYTel: +39. 06.98.358.472* *Cell: +39 331.2488468Fax: +39. 055.71.880.466Società del Gruppo SC31 ITALIA* ___ freebsd-virtualization@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-virtualization To unsubscribe, send any mail to freebsd-virtualization-unsubscr...@freebsd.org ___ freebsd-virtualization@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-virtualization To unsubscribe, send any mail to freebsd-virtualization-unsubscr...@freebsd.org -- *Andrea BrancatelliSchema 31 S.r.l. - Socio UnicoResponsabile ITROMA - FIRENZE - PALERMO ITALYTel: +39. 06.98.358.472* *Cell: +39 331.2488468Fax: +39. 055.71.880.466Società del Gruppo SC31 ITALIA* ___ freebsd-virtualization@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-virtualization To unsubscribe, send any mail to freebsd-virtualization-unsubscr...@freebsd.org
Re: BHyVe - ESXi comparison
On Tue, Jan 28, 2014 at 11:10 AM, Peter Grehan gre...@freebsd.org wrote: Hi Andrea, We did a very rough comparison betweend BHyVe and VMWare ESXi. Maybe you want to give it a read and let me know if I did write a bunch of sh!t :-) Looks good to me :) Thanks for running the tests. Would you be able to list the command options you used with bhyve when running these tests ? What I couldn't really understand (but that's something not related to bhyve or VMWare) is how a multiprocessor machine is slower than a singleprocessor machine in doing the compilation... any idea? Is hyper-threading enabled on your system ? If not, then with a host only having 2 CPUs and a 2 vCPU guest, there isn't as much opportunity to overlap host i/o threads with vCPU threads. Depends on how your setting up bhyve for example PetiteCloud limits it to no more then a 1 to 1 ratio of real to virtual cpu (after playing with something I will be posting about later today [me and Dee have the policy of no pronouncements] I am seriously thinking making this n vcpu to 1 real cpu... what do people think the best way to implement and the right ration for this is...) -- Aryeh M. Friedman, Lead Developer, http://www.PetiteCloud.org ___ freebsd-virtualization@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-virtualization To unsubscribe, send any mail to freebsd-virtualization-unsubscr...@freebsd.org
Re: BHyVe - ESXi comparison
It would be interesting to know to how much and what extent various fronends (openstack, cloudstack, petitecloud, etc.) effect performence... I suspect that even though bhyve is slower then VMWare that VMWare's front end is the cause and not VMWare itself for example I suspect that bhyve with petitecloud on top would be much faster then vmware or qemu with openstack or cloudstack. On Tue, Jan 28, 2014 at 6:18 AM, Andrea Brancatelli abrancate...@schema31.it wrote: Hello everybody. We did a very rough comparison betweend BHyVe and VMWare ESXi. Maybe you want to give it a read and let me know if I did write a bunch of sh!t :-) http://andrea.brancatelli.it/2014/01/28/freebsd-10-0-bhyve-vmware-esxi-5-5-comparison/ I must say I'm very pleased with BHyVe performances! Very good work!! Thanks for your time. -- *Andrea BrancatelliSchema 31 S.r.l. - Socio UnicoResponsabile ITROMA - FIRENZE - PALERMO ITALYTel: +39. 06.98.358.472* *Cell: +39 331.2488468Fax: +39. 055.71.880.466Società del Gruppo SC31 ITALIA* ___ freebsd-virtualization@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-virtualization To unsubscribe, send any mail to freebsd-virtualization-unsubscr...@freebsd.org -- Aryeh M. Friedman, Lead Developer, http://www.PetiteCloud.org ___ freebsd-virtualization@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-virtualization To unsubscribe, send any mail to freebsd-virtualization-unsubscr...@freebsd.org
Re: BHyVe - ESXi comparison
I have observed Ubuntu 12.04 LTS runs faster as a VM under bhyve the in does on bare metal (networking seems to be one of the key areas here as well as disk access) On Tue, Jan 28, 2014 at 9:57 AM, Lars Engels lars.eng...@0x20.net wrote: Am 2014-01-28 13:21, schrieb Andrea Brancatelli: Fixed, thanks. Could you also compare two instances of Linux inside bhyve and VMware? ___ freebsd-virtualization@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-virtualization To unsubscribe, send any mail to freebsd-virtualization- unsubscr...@freebsd.org -- Aryeh M. Friedman, Lead Developer, http://www.PetiteCloud.org ___ freebsd-virtualization@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-virtualization To unsubscribe, send any mail to freebsd-virtualization-unsubscr...@freebsd.org
Re: BHyVe - ESXi comparison
That's a lot of interesting input. Tomorrow we'll rearrange everything and redo all the testing. On Tue, Jan 28, 2014 at 9:50 PM, Peter Grehan gre...@freebsd.org wrote: Hi Andrea, unfortunately we've been a bit sloppy in tracking the time output because initially it was just an internal test, thus we don't have the details. No problems. We're setting up a new round of tests we'll run tomorrow and we'll track user/system/real in a more precise way; I will also publish a graph with the three stacked piles. Thanks, that'd be great. Some suggestions: if you're not already, I'd recommend using ahci-hd for disk images instead of virtio-blk. The AHCI emulation uses a thread for block i/o so won't block the VM on reads. Also, I'd recommend using a network login to the guest rather than running something from the console. The UART emulation in bhyve will result in a lot of VM exits, which can impact performance. later, Peter. -- *Andrea BrancatelliSchema 31 S.r.l. - Socio UnicoResponsabile ITROMA - FIRENZE - PALERMO ITALYTel: +39. 06.98.358.472* *Cell: +39 331.2488468Fax: +39. 055.71.880.466Società del Gruppo SC31 ITALIA* ___ freebsd-virtualization@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-virtualization To unsubscribe, send any mail to freebsd-virtualization-unsubscr...@freebsd.org
Re: BHyVe - ESXi comparison
Tomorrow we'll rearrange everything and redo all the testing. One more item: when running the test with 20 x 2-vCPU VMs, make sure that the -P option is being used. This forces bhyve to do a vmexit when a PAUSE instruction is hit e.g. when the locking code starts spinning. This gives the scheduler more opportunity to run something else rather than letting the VM go until it's quantum expires. later, Peter. ___ freebsd-virtualization@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-virtualization To unsubscribe, send any mail to freebsd-virtualization-unsubscr...@freebsd.org
BHyve - eESXi compatibility
Hi all, I follow the current BHyVe - ESXi comparison thread. To go further, I think one good reflexion topic is about the migration from one to another, back and forth. ESXi installation base is quite big, espacially on server side and with the people I meet. That is mostly french IT companies, having their cloud hosted by OVH (http://www.ovh.com/fr/dedicated-cloud/). In the company I'm working, we made an attempt to virtualize in-house with XenServer then export to OVH's ESXi when needed (back forth). The amount of work was too high in order to have a decent and reliable conversion, so that we gave up and stayed with the VMware tools for the moment. This conversion topic is really an important one. ___ freebsd-virtualization@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-virtualization To unsubscribe, send any mail to freebsd-virtualization-unsubscr...@freebsd.org
Re: Linux on BHyVe in 10.0-RELEASE
Where did you get the orignial starter script for this from was it from vmrun or the petitecloud sample script I posted... 2 reasons for asking: 1. It is turning out to be annoying inflexible in-terms of the values of various parameters (I can compensate for most of this in my personal playing with it to make it fit into petitecloud) 2. It uses a non-existent option (-I) Both are characterstics of petitecloud scripts but not so much of vmrun.sh (the main reason I do not recommend using petitecloud scripts except as nothing but barebone starter scripts if you use them to make your own). The reason for bringing all the above up is I am having a very hard time getting it to work and if it is from petitecloud can you please kindly walk me through how you went from my script to yours... the main issue I am attempting to over come is currently PC assumes that there is a single disk that represents the instance (a running instance is nothing more then loading it into RAM and wrapping a hyperv around it) [e.g. something like /vms/import/ubu.img] and that it contains *ALL* the data that is needed to boot (i.e. no other files are needed)... so far it appears (and I hope I am wrong) there is no way to force a linux instance into this model because you have to have the disk, the cd and the device map in the same dir but kept as separate files namely something like this: linux/ device.map disk.img cd.img seems to be the only model grub2-bhyve will be able to boot (symlinks at least on the surface seem to be a no go also...)... there are the following problems with this model though when doing mass VM's: * You have to copy the CD repeatedly to the boot dir (this is likely why openstack does not support cd based installs) * It makes for a really messy when attempting to make sure you completely nuke a vm when you delete it (currently pettiecloud does not delete the disk when the instance is deleted but this will be an option soon an likely the default) Of course the long term solution is unified off disk (vs. off loader) booting but until then any good work arounds? On Tue, Jan 28, 2014 at 6:23 AM, Markiyan Kushnir markiyan.kush...@gmail.com wrote: 2014-01-28 Aryeh Friedman aryeh.fried...@gmail.com: Seems like it is a processor motherboard combo thing (see other thread) ah, ok. good to know :) -- Markiyan. On Sun, Jan 26, 2014 at 9:26 AM, Markiyan Kushnir markiyan.kush...@gmail.com wrote: yes, once I posted it I realized that these sysctls are not relevant at this stage. There was a segfault, was that bhyvectl? May be it makes sense for someone (a bhyve dev) to inspect it? -- Markiyan. 2014-01-26 Aryeh Friedman aryeh.fried...@gmail.com: On Sun, Jan 26, 2014 at 8:46 AM, Markiyan Kushnir markiyan.kush...@gmail.com wrote: hmm... 15:45:test-bhyve# md5 ubuntu-12.04.3-server-amd64.iso MD5 (ubuntu-12.04.3-server-amd64.iso) = 2cbe868812a871242cdcdd8f2fd6feb9 # md5 ubuntu/ubuntu.iso MD5 (ubuntu/ubuntu.iso) = 2cbe868812a871242cdcdd8f2fd6feb9 doing the sysctl does no good -- Aryeh M. Friedman, Lead Developer, http://www.PetiteCloud.org -- Aryeh M. Friedman, Lead Developer, http://www.PetiteCloud.org -- Aryeh M. Friedman, Lead Developer, http://www.PetiteCloud.org ___ freebsd-virtualization@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-virtualization To unsubscribe, send any mail to freebsd-virtualization-unsubscr...@freebsd.org
Re: BHyVe - ESXi comparison
On 1/28/14 4:10 PM, Andrea Brancatelli wrote: That's a lot of interesting input. Do try the ahci-hd VirtIO and may I suggest you update BHyVe to the current bhyve? It is the first point on the FAQ: http://bhyve.org/faq/ Michael ___ freebsd-virtualization@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-virtualization To unsubscribe, send any mail to freebsd-virtualization-unsubscr...@freebsd.org
