Measuring ZFS configuration differences

classic Classic list List threaded Threaded
18 messages Options
Reply | Threaded
Open this post in threaded view
|

Measuring ZFS configuration differences

Dan Langille
Following up on the discussions during EuroBSDCon 2015 (Stockholm) during the FreeBSD Developer
Summit regarding various ZFS configuration settings, I write to start our implementation phase now that some
usual suspects have joined the list.

  re https://wiki.freebsd.org/201510DevSummit/Performance

I think the first order of business is granting access rights to the server (varm) in question:

  http://dan.langille.org/2015/07/19/varm/

During the workshop, mention was made of serial access.  I can arrange that.

The server has IPMI, however, my first thought:

1 - connect a USB-serial cable to varm & link that to another server in my rack.
2 - create a jail in that server and give it access to that serial connection
3 - redirect incoming port XYZ to that jail via a public-key-only ssh connection
4 - give people access

Any suggestions?


Dan Langille
http://langille.org/






signature.asc (990 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: Measuring ZFS configuration differences

Dan Langille

> On Oct 12, 2015, at 1:00 PM, Dan Langille <[hidden email]> wrote:
>
> Following up on the discussions during EuroBSDCon 2015 (Stockholm) during the FreeBSD Developer
> Summit regarding various ZFS configuration settings, I write to start our implementation phase now that some
> usual suspects have joined the list.
>
>  re https://wiki.freebsd.org/201510DevSummit/Performance
>
> I think the first order of business is granting access rights to the server (varm) in question:
>
>  http://dan.langille.org/2015/07/19/varm/
>
> During the workshop, mention was made of serial access.  I can arrange that.
>
> The server has IPMI, however, my first thought:
>
> 1 - connect a USB-serial cable to varm & link that to another server in my rack.
> 2 - create a jail in that server and give it access to that serial connection
> 3 - redirect incoming port XYZ to that jail via a public-key-only ssh connection
> 4 - give people access
>
> Any suggestions?
For the record, we are working with spinning HDD.  Donations of SSD (either physical or remote usage thereof)
are welcomed.  Our intent is to publish reproducible results which are not dependent upon your particular choice
of HDD (and hopefully, SSD).


Dan Langille
http://langille.org/






signature.asc (990 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: Measuring ZFS configuration differences

Marcelo Araujo-7
In reply to this post by Dan Langille
2015-10-13 1:00 GMT+08:00 Dan Langille <[hidden email]>:

>
>
> Any suggestions?
>
>
A list of "what we want to test" would be helpful. Maybe we can add it at
the same wiki page?
Also, would be nice assign each test to someone that wants to perform it.
As an example, I have my lab where I can make benchmarks in parallel with
your lab, if we assign tasks we won't have duplicated tests(in case we
don't want it).

I will update soon the information about my LAB.


Best.
--

--
Marcelo Araujo            (__)[hidden email]
\\\'',)http://www.FreeBSD.org <http://www.freebsd.org/>   \/  \ ^
Power To Server.         .\. /_)
_______________________________________________
[hidden email] mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-performance
To unsubscribe, send any mail to "[hidden email]"
Reply | Threaded
Open this post in threaded view
|

Re: Measuring ZFS configuration differences

Dan Langille

> On Oct 16, 2015, at 5:15 AM, Marcelo Araujo <[hidden email]> wrote:
>
> 2015-10-13 1:00 GMT+08:00 Dan Langille <[hidden email]>:
>
>>
>>
>> Any suggestions?
>>
>>
> A list of "what we want to test" would be helpful. Maybe we can add it at
> the same wiki page?
> Also, would be nice assign each test to someone that wants to perform it.
> As an example, I have my lab where I can make benchmarks in parallel with
> your lab, if we assign tasks we won't have duplicated tests(in case we
> don't want it).
>
> I will update soon the information about my LAB.

I think these are all fine ideas.

Running the same tests in more than one environment is good.  If our results can
be reproduced, it means our methods and conclusions are valid.


Dan Langille
http://langille.org/





_______________________________________________
[hidden email] mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-performance
To unsubscribe, send any mail to "[hidden email]"
Reply | Threaded
Open this post in threaded view
|

Re: Measuring ZFS configuration differences

Dan Langille
In reply to this post by Dan Langille
On Oct 12, 2015, at 1:00 PM, Dan Langille <[hidden email]> wrote:

>
> Following up on the discussions during EuroBSDCon 2015 (Stockholm) during the FreeBSD Developer
> Summit regarding various ZFS configuration settings, I write to start our implementation phase now that some
> usual suspects have joined the list.
>
>  re https://wiki.freebsd.org/201510DevSummit/Performance
>
> I think the first order of business is granting access rights to the server (varm) in question:
>
>  http://dan.langille.org/2015/07/19/varm/
>
> During the workshop, mention was made of serial access.  I can arrange that.
>
> The server has IPMI, however, my first thought:
>
> 1 - connect a USB-serial cable to varm & link that to another server in my rack.

Marcelo: At EuroBSDCon, was it you who mentioned a particular configuration for the test machine which made
it easy to configure and run tests?  Was it PXE booting or something?

> 2 - create a jail in that server and give it access to that serial connection
> 3 - redirect incoming port XYZ to that jail via a public-key-only ssh connection
> 4 - give people access
>
> Any suggestions?


Dan Langille
http://langille.org/

_______________________________________________
[hidden email] mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-performance
To unsubscribe, send any mail to "[hidden email]"
Reply | Threaded
Open this post in threaded view
|

Re: Measuring ZFS configuration differences

Marcelo Araujo-7
2015-11-12 6:34 GMT+08:00 Dan Langille <[hidden email]>:

> On Oct 12, 2015, at 1:00 PM, Dan Langille <[hidden email]> wrote:
> >
> > Following up on the discussions during EuroBSDCon 2015 (Stockholm)
> during the FreeBSD Developer
> > Summit regarding various ZFS configuration settings, I write to start
> our implementation phase now that some
> > usual suspects have joined the list.
> >
> >  re https://wiki.freebsd.org/201510DevSummit/Performance
> >
> > I think the first order of business is granting access rights to the
> server (varm) in question:
> >
> >  http://dan.langille.org/2015/07/19/varm/
> >
> > During the workshop, mention was made of serial access.  I can arrange
> that.
> >
> > The server has IPMI, however, my first thought:
> >
> > 1 - connect a USB-serial cable to varm & link that to another server in
> my rack.
>
> Marcelo: At EuroBSDCon, was it you who mentioned a particular
> configuration for the test machine which made
> it easy to configure and run tests?  Was it PXE booting or something?
>
> > 2 - create a jail in that server and give it access to that serial
> connection
> > 3 - redirect incoming port XYZ to that jail via a public-key-only ssh
> connection
> > 4 - give people access
> >
> > Any suggestions?
>
> —
> Dan Langille
> http://langille.org/
>
>
Hello Dan,

Yes, was me :)

I mention about zopkio test framework.
I gave a presentation last weekend at PyCon Hong Kong about it.

Here is my slides:
http://www.slideshare.net/araujobsd/functional-and-scale-performance-tests-using-zopkio

The good of Zopkio is, we can write tests at once and run it as much as we
want in different machines. Also Zopkio depends of Naarad, that can parse a
CSV file and create metrics and SLA over those metrics, plot graphs and so
on. Pretty nice tool!!!

I'm wondering if we could start to test something and maybe show it at
AsiaBSDCon and BSDCon(Canada) next year? What do you think?
What I need right now would be a list of tests that we want to perform as
well as what parameters we would like to take as metrics to compare.

Best Regards,


--

--
Marcelo Araujo            (__)[hidden email]
\\\'',)http://www.FreeBSD.org <http://www.freebsd.org/>   \/  \ ^
Power To Server.         .\. /_)
_______________________________________________
[hidden email] mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-performance
To unsubscribe, send any mail to "[hidden email]"
Reply | Threaded
Open this post in threaded view
|

Re: Measuring ZFS configuration differences

Allan Jude-9
On 2015-11-12 01:30, Marcelo Araujo wrote:

> 2015-11-12 6:34 GMT+08:00 Dan Langille <[hidden email]>:
>
>> On Oct 12, 2015, at 1:00 PM, Dan Langille <[hidden email]> wrote:
>>>
>>> Following up on the discussions during EuroBSDCon 2015 (Stockholm)
>> during the FreeBSD Developer
>>> Summit regarding various ZFS configuration settings, I write to start
>> our implementation phase now that some
>>> usual suspects have joined the list.
>>>
>>>  re https://wiki.freebsd.org/201510DevSummit/Performance
>>>
>>> I think the first order of business is granting access rights to the
>> server (varm) in question:
>>>
>>>  http://dan.langille.org/2015/07/19/varm/
>>>
>>> During the workshop, mention was made of serial access.  I can arrange
>> that.
>>>
>>> The server has IPMI, however, my first thought:
>>>
>>> 1 - connect a USB-serial cable to varm & link that to another server in
>> my rack.
>>
>> Marcelo: At EuroBSDCon, was it you who mentioned a particular
>> configuration for the test machine which made
>> it easy to configure and run tests?  Was it PXE booting or something?
>>
>>> 2 - create a jail in that server and give it access to that serial
>> connection
>>> 3 - redirect incoming port XYZ to that jail via a public-key-only ssh
>> connection
>>> 4 - give people access
>>>
>>> Any suggestions?
>>
>> —
>> Dan Langille
>> http://langille.org/
>>
>>
> Hello Dan,
>
> Yes, was me :)
>
> I mention about zopkio test framework.
> I gave a presentation last weekend at PyCon Hong Kong about it.
>
> Here is my slides:
> http://www.slideshare.net/araujobsd/functional-and-scale-performance-tests-using-zopkio
>
> The good of Zopkio is, we can write tests at once and run it as much as we
> want in different machines. Also Zopkio depends of Naarad, that can parse a
> CSV file and create metrics and SLA over those metrics, plot graphs and so
> on. Pretty nice tool!!!
>
> I'm wondering if we could start to test something and maybe show it at
> AsiaBSDCon and BSDCon(Canada) next year? What do you think?
> What I need right now would be a list of tests that we want to perform as
> well as what parameters we would like to take as metrics to compare.
>
> Best Regards,
>
>
Would this work for testing other things, like SSH performance? I am
looking at doing some SSH + HPN tests to prove that it is worth keeping.

--
Allan Jude


signature.asc (851 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: Measuring ZFS configuration differences

Marcelo Araujo-7
2015-11-12 14:38 GMT+08:00 Allan Jude <[hidden email]>:

> On 2015-11-12 01:30, Marcelo Araujo wrote:
> > 2015-11-12 6:34 GMT+08:00 Dan Langille <[hidden email]>:
> >
> >> On Oct 12, 2015, at 1:00 PM, Dan Langille <[hidden email]> wrote:
> >>>
> >>> Following up on the discussions during EuroBSDCon 2015 (Stockholm)
> >> during the FreeBSD Developer
> >>> Summit regarding various ZFS configuration settings, I write to start
> >> our implementation phase now that some
> >>> usual suspects have joined the list.
> >>>
> >>>  re https://wiki.freebsd.org/201510DevSummit/Performance
> >>>
> >>> I think the first order of business is granting access rights to the
> >> server (varm) in question:
> >>>
> >>>  http://dan.langille.org/2015/07/19/varm/
> >>>
> >>> During the workshop, mention was made of serial access.  I can arrange
> >> that.
> >>>
> >>> The server has IPMI, however, my first thought:
> >>>
> >>> 1 - connect a USB-serial cable to varm & link that to another server in
> >> my rack.
> >>
> >> Marcelo: At EuroBSDCon, was it you who mentioned a particular
> >> configuration for the test machine which made
> >> it easy to configure and run tests?  Was it PXE booting or something?
> >>
> >>> 2 - create a jail in that server and give it access to that serial
> >> connection
> >>> 3 - redirect incoming port XYZ to that jail via a public-key-only ssh
> >> connection
> >>> 4 - give people access
> >>>
> >>> Any suggestions?
> >>
> >> —
> >> Dan Langille
> >> http://langille.org/
> >>
> >>
> > Hello Dan,
> >
> > Yes, was me :)
> >
> > I mention about zopkio test framework.
> > I gave a presentation last weekend at PyCon Hong Kong about it.
> >
> > Here is my slides:
> >
> http://www.slideshare.net/araujobsd/functional-and-scale-performance-tests-using-zopkio
> >
> > The good of Zopkio is, we can write tests at once and run it as much as
> we
> > want in different machines. Also Zopkio depends of Naarad, that can
> parse a
> > CSV file and create metrics and SLA over those metrics, plot graphs and
> so
> > on. Pretty nice tool!!!
> >
> > I'm wondering if we could start to test something and maybe show it at
> > AsiaBSDCon and BSDCon(Canada) next year? What do you think?
> > What I need right now would be a list of tests that we want to perform as
> > well as what parameters we would like to take as metrics to compare.
> >
> > Best Regards,
> >
> >
>
> Would this work for testing other things, like SSH performance? I am
> looking at doing some SSH + HPN tests to prove that it is worth keeping.
>
> --
> Allan Jude
>
>
It would depends what kind of performance you want to test!
If the metrics you can output to a CSV file, the answer would be yes, it
can!


Best,
--

--
Marcelo Araujo            (__)[hidden email]
\\\'',)http://www.FreeBSD.org <http://www.freebsd.org/>   \/  \ ^
Power To Server.         .\. /_)
_______________________________________________
[hidden email] mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-performance
To unsubscribe, send any mail to "[hidden email]"
Reply | Threaded
Open this post in threaded view
|

Re: Measuring ZFS configuration differences

Dan Langille
In reply to this post by Marcelo Araujo-7

> On Nov 12, 2015, at 1:30 AM, Marcelo Araujo <[hidden email]> wrote:
>
>
>
> 2015-11-12 6:34 GMT+08:00 Dan Langille <[hidden email] <mailto:[hidden email]>>:
> On Oct 12, 2015, at 1:00 PM, Dan Langille <[hidden email] <mailto:[hidden email]>> wrote:
> >
> > Following up on the discussions during EuroBSDCon 2015 (Stockholm) during the FreeBSD Developer
> > Summit regarding various ZFS configuration settings, I write to start our implementation phase now that some
> > usual suspects have joined the list.
> >
> >  re https://wiki.freebsd.org/201510DevSummit/Performance <https://wiki.freebsd.org/201510DevSummit/Performance>
> >
> > I think the first order of business is granting access rights to the server (varm) in question:
> >
> >  http://dan.langille.org/2015/07/19/varm/ <http://dan.langille.org/2015/07/19/varm/>
> >
> > During the workshop, mention was made of serial access.  I can arrange that.
> >
> > The server has IPMI, however, my first thought:
> >
> > 1 - connect a USB-serial cable to varm & link that to another server in my rack.
>
> Marcelo: At EuroBSDCon, was it you who mentioned a particular configuration for the test machine which made
> it easy to configure and run tests?  Was it PXE booting or something?
>
> > 2 - create a jail in that server and give it access to that serial connection
> > 3 - redirect incoming port XYZ to that jail via a public-key-only ssh connection
> > 4 - give people access
> >
> > Any suggestions?
>
> —
> Dan Langille
> http://langille.org/ <http://langille.org/>
>
>
> Hello Dan,
>
> Yes, was me :)
>
> I mention about zopkio test framework.
> I gave a presentation last weekend at PyCon Hong Kong about it.
>
> Here is my slides: http://www.slideshare.net/araujobsd/functional-and-scale-performance-tests-using-zopkio <http://www.slideshare.net/araujobsd/functional-and-scale-performance-tests-using-zopkio>
>
> The good of Zopkio is, we can write tests at once and run it as much as we want in different machines. Also Zopkio depends of Naarad, that can parse a CSV file and create metrics and SLA over those metrics, plot graphs and so on. Pretty nice tool!!!
>
> I'm wondering if we could start to test something and maybe show it at AsiaBSDCon and BSDCon(Canada) next year? What do you think?
> What I need right now would be a list of tests that we want to perform as well as what parameters we would like to take as metrics to compare.\
For tests, we can start with this list: https://github.com/dlangille/zfs_benchmarks/issues <https://github.com/dlangille/zfs_benchmarks/issues>

We can start as soon as I figure out how to provide access to the testers.  See above re serial connection.

I want to provide access, but I want to keep access restricted to only
 this box and not to the rest of my home LAN.  I plan to do this via a VLAN.

I could fire up a Rasperberry Pi and allow ssh into that.  Will that be enough
power for what you need to do?


Dan Langille
http://langille <http://langille/>.org/






signature.asc (990 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: Measuring ZFS configuration differences

Marcelo Araujo-7
2015-11-18 3:14 GMT+08:00 Dan Langille <[hidden email]>:

>
> On Nov 12, 2015, at 1:30 AM, Marcelo Araujo <[hidden email]>
> wrote:
>
>
>
> 2015-11-12 6:34 GMT+08:00 Dan Langille <[hidden email]>:
>
>> On Oct 12, 2015, at 1:00 PM, Dan Langille <[hidden email]> wrote:
>> >
>> > Following up on the discussions during EuroBSDCon 2015 (Stockholm)
>> during the FreeBSD Developer
>> > Summit regarding various ZFS configuration settings, I write to start
>> our implementation phase now that some
>> > usual suspects have joined the list.
>> >
>> >  re https://wiki.freebsd.org/201510DevSummit/Performance
>> >
>> > I think the first order of business is granting access rights to the
>> server (varm) in question:
>> >
>> >  http://dan.langille.org/2015/07/19/varm/
>> >
>> > During the workshop, mention was made of serial access.  I can arrange
>> that.
>> >
>> > The server has IPMI, however, my first thought:
>> >
>> > 1 - connect a USB-serial cable to varm & link that to another server in
>> my rack.
>>
>> Marcelo: At EuroBSDCon, was it you who mentioned a particular
>> configuration for the test machine which made
>> it easy to configure and run tests?  Was it PXE booting or something?
>>
>> > 2 - create a jail in that server and give it access to that serial
>> connection
>> > 3 - redirect incoming port XYZ to that jail via a public-key-only ssh
>> connection
>> > 4 - give people access
>> >
>> > Any suggestions?
>>
>> —
>> Dan Langille
>> http://langille.org/
>>
>>
> Hello Dan,
>
> Yes, was me :)
>
> I mention about zopkio test framework.
> I gave a presentation last weekend at PyCon Hong Kong about it.
>
> Here is my slides:
> http://www.slideshare.net/araujobsd/functional-and-scale-performance-tests-using-zopkio
>
> The good of Zopkio is, we can write tests at once and run it as much as we
> want in different machines. Also Zopkio depends of Naarad, that can parse a
> CSV file and create metrics and SLA over those metrics, plot graphs and so
> on. Pretty nice tool!!!
>
> I'm wondering if we could start to test something and maybe show it at
> AsiaBSDCon and BSDCon(Canada) next year? What do you think?
> What I need right now would be a list of tests that we want to perform as
> well as what parameters we would like to take as metrics to compare.\
>
>
> For tests, we can start with this list:
> https://github.com/dlangille/zfs_benchmarks/issues
>
> We can start as soon as I figure out how to provide access to the
> testers.  See above re serial connection.
>
> I want to provide access, but I want to keep access restricted to only
>  this box and not to the rest of my home LAN.  I plan to do this via a
> VLAN.
>
> I could fire up a Rasperberry Pi and allow ssh into that.  Will that be
> enough
> power for what you need to do?
>
>
First of all, thanks to share the tests cases.

If I use zopkio, the best would be access SSH direct to the target machine
where I need to run the tests. For zopkio, I need to have my SSH KEY on the
target machine.

As I don't know your network, maybe what you could do is: Via RasperBerry,
forward the SSH to the target machine, I will pass-through via your
RasperBerry where you can control the access for the rest of your LAN.

Another approach could be, two different subnets and a firewall. Or as you
said, VLANS.


Best,
--

--
Marcelo Araujo            (__)[hidden email]
\\\'',)http://www.FreeBSD.org <http://www.freebsd.org/>   \/  \ ^
Power To Server.         .\. /_)
_______________________________________________
[hidden email] mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-performance
To unsubscribe, send any mail to "[hidden email]"
Reply | Threaded
Open this post in threaded view
|

Re: Measuring ZFS configuration differences

Dan Langille

> On Nov 17, 2015, at 8:46 PM, Marcelo Araujo <[hidden email]> wrote:
>
> 2015-11-18 3:14 GMT+08:00 Dan Langille <[hidden email]>:
>
>>
>> On Nov 12, 2015, at 1:30 AM, Marcelo Araujo <[hidden email]>
>> wrote:
>>
>>
>>
>> 2015-11-12 6:34 GMT+08:00 Dan Langille <[hidden email]>:
>>
>>> On Oct 12, 2015, at 1:00 PM, Dan Langille <[hidden email]> wrote:
>>>>
>>>> Following up on the discussions during EuroBSDCon 2015 (Stockholm)
>>> during the FreeBSD Developer
>>>> Summit regarding various ZFS configuration settings, I write to start
>>> our implementation phase now that some
>>>> usual suspects have joined the list.
>>>>
>>>> re https://wiki.freebsd.org/201510DevSummit/Performance
>>>>
>>>> I think the first order of business is granting access rights to the
>>> server (varm) in question:
>>>>
>>>> http://dan.langille.org/2015/07/19/varm/
>>>>
>>>> During the workshop, mention was made of serial access.  I can arrange
>>> that.
>>>>
>>>> The server has IPMI, however, my first thought:
>>>>
>>>> 1 - connect a USB-serial cable to varm & link that to another server in
>>> my rack.
>>>
>>> Marcelo: At EuroBSDCon, was it you who mentioned a particular
>>> configuration for the test machine which made
>>> it easy to configure and run tests?  Was it PXE booting or something?
>>>
>>>> 2 - create a jail in that server and give it access to that serial
>>> connection
>>>> 3 - redirect incoming port XYZ to that jail via a public-key-only ssh
>>> connection
>>>> 4 - give people access
>>>>
>>>> Any suggestions?
>>>
>>> —
>>> Dan Langille
>>> http://langille.org/
>>>
>>>
>> Hello Dan,
>>
>> Yes, was me :)
>>
>> I mention about zopkio test framework.
>> I gave a presentation last weekend at PyCon Hong Kong about it.
>>
>> Here is my slides:
>> http://www.slideshare.net/araujobsd/functional-and-scale-performance-tests-using-zopkio
>>
>> The good of Zopkio is, we can write tests at once and run it as much as we
>> want in different machines. Also Zopkio depends of Naarad, that can parse a
>> CSV file and create metrics and SLA over those metrics, plot graphs and so
>> on. Pretty nice tool!!!
>>
>> I'm wondering if we could start to test something and maybe show it at
>> AsiaBSDCon and BSDCon(Canada) next year? What do you think?
>> What I need right now would be a list of tests that we want to perform as
>> well as what parameters we would like to take as metrics to compare.\
>>
>>
>> For tests, we can start with this list:
>> https://github.com/dlangille/zfs_benchmarks/issues
>>
>> We can start as soon as I figure out how to provide access to the
>> testers.  See above re serial connection.
>>
>> I want to provide access, but I want to keep access restricted to only
>> this box and not to the rest of my home LAN.  I plan to do this via a
>> VLAN.
>>
>> I could fire up a Rasperberry Pi and allow ssh into that.  Will that be
>> enough
>> power for what you need to do?
>>
>>
> First of all, thanks to share the tests cases.
>
> If I use zopkio, the best would be access SSH direct to the target machine
> where I need to run the tests. For zopkio, I need to have my SSH KEY on the
> target machine.

I am OK with this.

> As I don't know your network, maybe what you could do is: Via RasperBerry,
> forward the SSH to the target machine, I will pass-through via your
> RasperBerry where you can control the access for the rest of your LAN.
>
> Another approach could be, two different subnets and a firewall. Or as you
> said, VLANS.

I will be doing VLANS, which have yet to be set up.

The target system will have ZFS pools can be configured for different tests (i.e. raidz2 vs raidz3).
This will involve gpart etc because the drives & pools will need to be 'wiped' between different test
runs.

I seem to recall someone suggesting PXE boot and configuring the system remotely.  Does anyone
recall that?  That aspect of the discussion was not recorded: https://wiki.freebsd.org/201510DevSummit/Performance


Dan Langille
http://langille.org/





_______________________________________________
[hidden email] mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-performance
To unsubscribe, send any mail to "[hidden email]"
Reply | Threaded
Open this post in threaded view
|

Re: Measuring ZFS configuration differences

Marcelo Araujo-7
2015-11-18 10:23 GMT+08:00 Dan Langille <[hidden email]>:

>
> > On Nov 17, 2015, at 8:46 PM, Marcelo Araujo <[hidden email]>
> wrote:
> >
> > 2015-11-18 3:14 GMT+08:00 Dan Langille <[hidden email]>:
> >
> >>
> >> On Nov 12, 2015, at 1:30 AM, Marcelo Araujo <[hidden email]>
> >> wrote:
> >>
> >>
> >>
> >> 2015-11-12 6:34 GMT+08:00 Dan Langille <[hidden email]>:
> >>
> >>> On Oct 12, 2015, at 1:00 PM, Dan Langille <[hidden email]> wrote:
> >>>>
> >>>> Following up on the discussions during EuroBSDCon 2015 (Stockholm)
> >>> during the FreeBSD Developer
> >>>> Summit regarding various ZFS configuration settings, I write to start
> >>> our implementation phase now that some
> >>>> usual suspects have joined the list.
> >>>>
> >>>> re https://wiki.freebsd.org/201510DevSummit/Performance
> >>>>
> >>>> I think the first order of business is granting access rights to the
> >>> server (varm) in question:
> >>>>
> >>>> http://dan.langille.org/2015/07/19/varm/
> >>>>
> >>>> During the workshop, mention was made of serial access.  I can arrange
> >>> that.
> >>>>
> >>>> The server has IPMI, however, my first thought:
> >>>>
> >>>> 1 - connect a USB-serial cable to varm & link that to another server
> in
> >>> my rack.
> >>>
> >>> Marcelo: At EuroBSDCon, was it you who mentioned a particular
> >>> configuration for the test machine which made
> >>> it easy to configure and run tests?  Was it PXE booting or something?
> >>>
> >>>> 2 - create a jail in that server and give it access to that serial
> >>> connection
> >>>> 3 - redirect incoming port XYZ to that jail via a public-key-only ssh
> >>> connection
> >>>> 4 - give people access
> >>>>
> >>>> Any suggestions?
> >>>
> >>> —
> >>> Dan Langille
> >>> http://langille.org/
> >>>
> >>>
> >> Hello Dan,
> >>
> >> Yes, was me :)
> >>
> >> I mention about zopkio test framework.
> >> I gave a presentation last weekend at PyCon Hong Kong about it.
> >>
> >> Here is my slides:
> >>
> http://www.slideshare.net/araujobsd/functional-and-scale-performance-tests-using-zopkio
> >>
> >> The good of Zopkio is, we can write tests at once and run it as much as
> we
> >> want in different machines. Also Zopkio depends of Naarad, that can
> parse a
> >> CSV file and create metrics and SLA over those metrics, plot graphs and
> so
> >> on. Pretty nice tool!!!
> >>
> >> I'm wondering if we could start to test something and maybe show it at
> >> AsiaBSDCon and BSDCon(Canada) next year? What do you think?
> >> What I need right now would be a list of tests that we want to perform
> as
> >> well as what parameters we would like to take as metrics to compare.\
> >>
> >>
> >> For tests, we can start with this list:
> >> https://github.com/dlangille/zfs_benchmarks/issues
> >>
> >> We can start as soon as I figure out how to provide access to the
> >> testers.  See above re serial connection.
> >>
> >> I want to provide access, but I want to keep access restricted to only
> >> this box and not to the rest of my home LAN.  I plan to do this via a
> >> VLAN.
> >>
> >> I could fire up a Rasperberry Pi and allow ssh into that.  Will that be
> >> enough
> >> power for what you need to do?
> >>
> >>
> > First of all, thanks to share the tests cases.
> >
> > If I use zopkio, the best would be access SSH direct to the target
> machine
> > where I need to run the tests. For zopkio, I need to have my SSH KEY on
> the
> > target machine.
>
> I am OK with this.
>
> > As I don't know your network, maybe what you could do is: Via
> RasperBerry,
> > forward the SSH to the target machine, I will pass-through via your
> > RasperBerry where you can control the access for the rest of your LAN.
> >
> > Another approach could be, two different subnets and a firewall. Or as
> you
> > said, VLANS.
>
> I will be doing VLANS, which have yet to be set up.
>
> The target system will have ZFS pools can be configured for different
> tests (i.e. raidz2 vs raidz3).
> This will involve gpart etc because the drives & pools will need to be
> 'wiped' between different test
> runs.
>
> I seem to recall someone suggesting PXE boot and configuring the system
> remotely.  Does anyone
> recall that?  That aspect of the discussion was not recorded:
> https://wiki.freebsd.org/201510DevSummit/Performance
>
>
Bapt@ mentioned that, this is the way how we are doing in another project.
But in my point of view, it is not a must for our case!

The PXE wold be good if we try to test different of OS flavors, or build
different images.


Best,
--

--
Marcelo Araujo            (__)[hidden email]
\\\'',)http://www.FreeBSD.org <http://www.freebsd.org/>   \/  \ ^
Power To Server.         .\. /_)
_______________________________________________
[hidden email] mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-performance
To unsubscribe, send any mail to "[hidden email]"
Reply | Threaded
Open this post in threaded view
|

Re: Measuring ZFS configuration differences

Dan Langille

> On Nov 17, 2015, at 9:31 PM, Marcelo Araujo <[hidden email]> wrote:
>
> 2015-11-18 10:23 GMT+08:00 Dan Langille <[hidden email]>:
>
>>
>>> On Nov 17, 2015, at 8:46 PM, Marcelo Araujo <[hidden email]>
>> wrote:
>>>
>>> 2015-11-18 3:14 GMT+08:00 Dan Langille <[hidden email]>:
>>>
>>>>
>>>> On Nov 12, 2015, at 1:30 AM, Marcelo Araujo <[hidden email]>
>>>> wrote:
>>>>
>>>>
>>>>
>>>> 2015-11-12 6:34 GMT+08:00 Dan Langille <[hidden email]>:
>>>>
>>>>> On Oct 12, 2015, at 1:00 PM, Dan Langille <[hidden email]> wrote:
>>>>>>
>>>>>> Following up on the discussions during EuroBSDCon 2015 (Stockholm)
>>>>> during the FreeBSD Developer
>>>>>> Summit regarding various ZFS configuration settings, I write to start
>>>>> our implementation phase now that some
>>>>>> usual suspects have joined the list.
>>>>>>
>>>>>> re https://wiki.freebsd.org/201510DevSummit/Performance
>>>>>>
>>>>>> I think the first order of business is granting access rights to the
>>>>> server (varm) in question:
>>>>>>
>>>>>> http://dan.langille.org/2015/07/19/varm/
>>>>>>
>>>>>> During the workshop, mention was made of serial access.  I can arrange
>>>>> that.
>>>>>>
>>>>>> The server has IPMI, however, my first thought:
>>>>>>
>>>>>> 1 - connect a USB-serial cable to varm & link that to another server
>> in
>>>>> my rack.
>>>>>
>>>>> Marcelo: At EuroBSDCon, was it you who mentioned a particular
>>>>> configuration for the test machine which made
>>>>> it easy to configure and run tests?  Was it PXE booting or something?
>>>>>
>>>>>> 2 - create a jail in that server and give it access to that serial
>>>>> connection
>>>>>> 3 - redirect incoming port XYZ to that jail via a public-key-only ssh
>>>>> connection
>>>>>> 4 - give people access
>>>>>>
>>>>>> Any suggestions?
>>>>>
>>>>> —
>>>>> Dan Langille
>>>>> http://langille.org/
>>>>>
>>>>>
>>>> Hello Dan,
>>>>
>>>> Yes, was me :)
>>>>
>>>> I mention about zopkio test framework.
>>>> I gave a presentation last weekend at PyCon Hong Kong about it.
>>>>
>>>> Here is my slides:
>>>>
>> http://www.slideshare.net/araujobsd/functional-and-scale-performance-tests-using-zopkio
>>>>
>>>> The good of Zopkio is, we can write tests at once and run it as much as
>> we
>>>> want in different machines. Also Zopkio depends of Naarad, that can
>> parse a
>>>> CSV file and create metrics and SLA over those metrics, plot graphs and
>> so
>>>> on. Pretty nice tool!!!
>>>>
>>>> I'm wondering if we could start to test something and maybe show it at
>>>> AsiaBSDCon and BSDCon(Canada) next year? What do you think?
>>>> What I need right now would be a list of tests that we want to perform
>> as
>>>> well as what parameters we would like to take as metrics to compare.\
>>>>
>>>>
>>>> For tests, we can start with this list:
>>>> https://github.com/dlangille/zfs_benchmarks/issues
>>>>
>>>> We can start as soon as I figure out how to provide access to the
>>>> testers.  See above re serial connection.
>>>>
>>>> I want to provide access, but I want to keep access restricted to only
>>>> this box and not to the rest of my home LAN.  I plan to do this via a
>>>> VLAN.
>>>>
>>>> I could fire up a Rasperberry Pi and allow ssh into that.  Will that be
>>>> enough
>>>> power for what you need to do?
>>>>
>>>>
>>> First of all, thanks to share the tests cases.
>>>
>>> If I use zopkio, the best would be access SSH direct to the target
>> machine
>>> where I need to run the tests. For zopkio, I need to have my SSH KEY on
>> the
>>> target machine.
>>
>> I am OK with this.
>>
>>> As I don't know your network, maybe what you could do is: Via
>> RasperBerry,
>>> forward the SSH to the target machine, I will pass-through via your
>>> RasperBerry where you can control the access for the rest of your LAN.
>>>
>>> Another approach could be, two different subnets and a firewall. Or as
>> you
>>> said, VLANS.
>>
>> I will be doing VLANS, which have yet to be set up.
>>
>> The target system will have ZFS pools can be configured for different
>> tests (i.e. raidz2 vs raidz3).
>> This will involve gpart etc because the drives & pools will need to be
>> 'wiped' between different test
>> runs.
>>
>> I seem to recall someone suggesting PXE boot and configuring the system
>> remotely.  Does anyone
>> recall that?  That aspect of the discussion was not recorded:
>> https://wiki.freebsd.org/201510DevSummit/Performance
>>
>>
> Bapt@ mentioned that, this is the way how we are doing in another project.
> But in my point of view, it is not a must for our case!
>
> The PXE wold be good if we try to test different of OS flavors, or build
> different images.

OK.  The only thing holding us back is:

- adding the air filters to the case
- moving to the new switch with the new VLANs

It's now a matter of time.


Dan Langille
http://langille.org/





_______________________________________________
[hidden email] mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-performance
To unsubscribe, send any mail to "[hidden email]"
Reply | Threaded
Open this post in threaded view
|

Re: Measuring ZFS configuration differences

Tom Curry
On Wed, Nov 18, 2015 at 4:12 PM, Dan Langille <[hidden email]> wrote:

>
> > On Nov 17, 2015, at 9:31 PM, Marcelo Araujo <[hidden email]>
> wrote:
> >
> > 2015-11-18 10:23 GMT+08:00 Dan Langille <[hidden email]>:
> >
> >>
> >>> On Nov 17, 2015, at 8:46 PM, Marcelo Araujo <[hidden email]>
> >> wrote:
> >>>
> >>> 2015-11-18 3:14 GMT+08:00 Dan Langille <[hidden email]>:
> >>>
> >>>>
> >>>> On Nov 12, 2015, at 1:30 AM, Marcelo Araujo <[hidden email]>
> >>>> wrote:
> >>>>
> >>>>
> >>>>
> >>>> 2015-11-12 6:34 GMT+08:00 Dan Langille <[hidden email]>:
> >>>>
> >>>>> On Oct 12, 2015, at 1:00 PM, Dan Langille <[hidden email]> wrote:
> >>>>>>
> >>>>>> Following up on the discussions during EuroBSDCon 2015 (Stockholm)
> >>>>> during the FreeBSD Developer
> >>>>>> Summit regarding various ZFS configuration settings, I write to
> start
> >>>>> our implementation phase now that some
> >>>>>> usual suspects have joined the list.
> >>>>>>
> >>>>>> re https://wiki.freebsd.org/201510DevSummit/Performance
> >>>>>>
> >>>>>> I think the first order of business is granting access rights to the
> >>>>> server (varm) in question:
> >>>>>>
> >>>>>> http://dan.langille.org/2015/07/19/varm/
> >>>>>>
> >>>>>> During the workshop, mention was made of serial access.  I can
> arrange
> >>>>> that.
> >>>>>>
> >>>>>> The server has IPMI, however, my first thought:
> >>>>>>
> >>>>>> 1 - connect a USB-serial cable to varm & link that to another server
> >> in
> >>>>> my rack.
> >>>>>
> >>>>> Marcelo: At EuroBSDCon, was it you who mentioned a particular
> >>>>> configuration for the test machine which made
> >>>>> it easy to configure and run tests?  Was it PXE booting or something?
> >>>>>
> >>>>>> 2 - create a jail in that server and give it access to that serial
> >>>>> connection
> >>>>>> 3 - redirect incoming port XYZ to that jail via a public-key-only
> ssh
> >>>>> connection
> >>>>>> 4 - give people access
> >>>>>>
> >>>>>> Any suggestions?
> >>>>>
> >>>>> —
> >>>>> Dan Langille
> >>>>> http://langille.org/
> >>>>>
> >>>>>
> >>>> Hello Dan,
> >>>>
> >>>> Yes, was me :)
> >>>>
> >>>> I mention about zopkio test framework.
> >>>> I gave a presentation last weekend at PyCon Hong Kong about it.
> >>>>
> >>>> Here is my slides:
> >>>>
> >>
> http://www.slideshare.net/araujobsd/functional-and-scale-performance-tests-using-zopkio
> >>>>
> >>>> The good of Zopkio is, we can write tests at once and run it as much
> as
> >> we
> >>>> want in different machines. Also Zopkio depends of Naarad, that can
> >> parse a
> >>>> CSV file and create metrics and SLA over those metrics, plot graphs
> and
> >> so
> >>>> on. Pretty nice tool!!!
> >>>>
> >>>> I'm wondering if we could start to test something and maybe show it at
> >>>> AsiaBSDCon and BSDCon(Canada) next year? What do you think?
> >>>> What I need right now would be a list of tests that we want to perform
> >> as
> >>>> well as what parameters we would like to take as metrics to compare.\
> >>>>
> >>>>
> >>>> For tests, we can start with this list:
> >>>> https://github.com/dlangille/zfs_benchmarks/issues
> >>>>
> >>>> We can start as soon as I figure out how to provide access to the
> >>>> testers.  See above re serial connection.
> >>>>
> >>>> I want to provide access, but I want to keep access restricted to only
> >>>> this box and not to the rest of my home LAN.  I plan to do this via a
> >>>> VLAN.
> >>>>
> >>>> I could fire up a Rasperberry Pi and allow ssh into that.  Will that
> be
> >>>> enough
> >>>> power for what you need to do?
> >>>>
> >>>>
> >>> First of all, thanks to share the tests cases.
> >>>
> >>> If I use zopkio, the best would be access SSH direct to the target
> >> machine
> >>> where I need to run the tests. For zopkio, I need to have my SSH KEY on
> >> the
> >>> target machine.
> >>
> >> I am OK with this.
> >>
> >>> As I don't know your network, maybe what you could do is: Via
> >> RasperBerry,
> >>> forward the SSH to the target machine, I will pass-through via your
> >>> RasperBerry where you can control the access for the rest of your LAN.
> >>>
> >>> Another approach could be, two different subnets and a firewall. Or as
> >> you
> >>> said, VLANS.
> >>
> >> I will be doing VLANS, which have yet to be set up.
> >>
> >> The target system will have ZFS pools can be configured for different
> >> tests (i.e. raidz2 vs raidz3).
> >> This will involve gpart etc because the drives & pools will need to be
> >> 'wiped' between different test
> >> runs.
> >>
> >> I seem to recall someone suggesting PXE boot and configuring the system
> >> remotely.  Does anyone
> >> recall that?  That aspect of the discussion was not recorded:
> >> https://wiki.freebsd.org/201510DevSummit/Performance
> >>
> >>
> > Bapt@ mentioned that, this is the way how we are doing in another
> project.
> > But in my point of view, it is not a must for our case!
> >
> > The PXE wold be good if we try to test different of OS flavors, or build
> > different images.
>
> OK.  The only thing holding us back is:
>
> - adding the air filters to the case
> - moving to the new switch with the new VLANs
>
> It's now a matter of time.
>
> —
> Dan Langille
> http://langille.org/
>
>
>
> _______________________________________________
> [hidden email] mailing list
> https://lists.freebsd.org/mailman/listinfo/freebsd-performance
> To unsubscribe, send any mail to "
> [hidden email]"
>


Dan,

I don't know if you have already cracked the serial access nut, but every
supermicro motherboard I have owned has supported serial over lan. I
personally have never used it, I have just noticed it from time to time.
Here is a good resource on setting it up
http://serverfault.com/questions/574351/serial-over-lan-on-freebsd-10-0-with-supermicro-x9-scm-f
(note: it's for an X9 motherboard but it should work for your X10, see the
comment at the very end)

Also, FreeBSD, ZFS and performance are three of my most favorite words. Is
there any way I can participate in this adventure?
_______________________________________________
[hidden email] mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-performance
To unsubscribe, send any mail to "[hidden email]"
Reply | Threaded
Open this post in threaded view
|

Re: Measuring ZFS configuration differences

Dan Langille

> On Nov 18, 2015, at 5:47 PM, Tom Curry <[hidden email]> wrote:
>
> On Wed, Nov 18, 2015 at 4:12 PM, Dan Langille <[hidden email]> wrote:
>
>>
>>> On Nov 17, 2015, at 9:31 PM, Marcelo Araujo <[hidden email]>
>> wrote:
>>>
>>> 2015-11-18 10:23 GMT+08:00 Dan Langille <[hidden email]>:
>>>
>>>>
>>>>> On Nov 17, 2015, at 8:46 PM, Marcelo Araujo <[hidden email]>
>>>> wrote:
>>>>>
>>>>> 2015-11-18 3:14 GMT+08:00 Dan Langille <[hidden email]>:
>>>>>
>>>>>>
>>>>>> On Nov 12, 2015, at 1:30 AM, Marcelo Araujo <[hidden email]>
>>>>>> wrote:
>>>>>>
>>>>>>
>>>>>>
>>>>>> 2015-11-12 6:34 GMT+08:00 Dan Langille <[hidden email]>:
>>>>>>
>>>>>>> On Oct 12, 2015, at 1:00 PM, Dan Langille <[hidden email]> wrote:
>>>>>>>>
>>>>>>>> Following up on the discussions during EuroBSDCon 2015 (Stockholm)
>>>>>>> during the FreeBSD Developer
>>>>>>>> Summit regarding various ZFS configuration settings, I write to
>> start
>>>>>>> our implementation phase now that some
>>>>>>>> usual suspects have joined the list.
>>>>>>>>
>>>>>>>> re https://wiki.freebsd.org/201510DevSummit/Performance
>>>>>>>>
>>>>>>>> I think the first order of business is granting access rights to the
>>>>>>> server (varm) in question:
>>>>>>>>
>>>>>>>> http://dan.langille.org/2015/07/19/varm/
>>>>>>>>
>>>>>>>> During the workshop, mention was made of serial access.  I can
>> arrange
>>>>>>> that.
>>>>>>>>
>>>>>>>> The server has IPMI, however, my first thought:
>>>>>>>>
>>>>>>>> 1 - connect a USB-serial cable to varm & link that to another server
>>>> in
>>>>>>> my rack.
>>>>>>>
>>>>>>> Marcelo: At EuroBSDCon, was it you who mentioned a particular
>>>>>>> configuration for the test machine which made
>>>>>>> it easy to configure and run tests?  Was it PXE booting or something?
>>>>>>>
>>>>>>>> 2 - create a jail in that server and give it access to that serial
>>>>>>> connection
>>>>>>>> 3 - redirect incoming port XYZ to that jail via a public-key-only
>> ssh
>>>>>>> connection
>>>>>>>> 4 - give people access
>>>>>>>>
>>>>>>>> Any suggestions?
>>>>>>>
>>>>>>> —
>>>>>>> Dan Langille
>>>>>>> http://langille.org/
>>>>>>>
>>>>>>>
>>>>>> Hello Dan,
>>>>>>
>>>>>> Yes, was me :)
>>>>>>
>>>>>> I mention about zopkio test framework.
>>>>>> I gave a presentation last weekend at PyCon Hong Kong about it.
>>>>>>
>>>>>> Here is my slides:
>>>>>>
>>>>
>> http://www.slideshare.net/araujobsd/functional-and-scale-performance-tests-using-zopkio
>>>>>>
>>>>>> The good of Zopkio is, we can write tests at once and run it as much
>> as
>>>> we
>>>>>> want in different machines. Also Zopkio depends of Naarad, that can
>>>> parse a
>>>>>> CSV file and create metrics and SLA over those metrics, plot graphs
>> and
>>>> so
>>>>>> on. Pretty nice tool!!!
>>>>>>
>>>>>> I'm wondering if we could start to test something and maybe show it at
>>>>>> AsiaBSDCon and BSDCon(Canada) next year? What do you think?
>>>>>> What I need right now would be a list of tests that we want to perform
>>>> as
>>>>>> well as what parameters we would like to take as metrics to compare.\
>>>>>>
>>>>>>
>>>>>> For tests, we can start with this list:
>>>>>> https://github.com/dlangille/zfs_benchmarks/issues
>>>>>>
>>>>>> We can start as soon as I figure out how to provide access to the
>>>>>> testers.  See above re serial connection.
>>>>>>
>>>>>> I want to provide access, but I want to keep access restricted to only
>>>>>> this box and not to the rest of my home LAN.  I plan to do this via a
>>>>>> VLAN.
>>>>>>
>>>>>> I could fire up a Rasperberry Pi and allow ssh into that.  Will that
>> be
>>>>>> enough
>>>>>> power for what you need to do?
>>>>>>
>>>>>>
>>>>> First of all, thanks to share the tests cases.
>>>>>
>>>>> If I use zopkio, the best would be access SSH direct to the target
>>>> machine
>>>>> where I need to run the tests. For zopkio, I need to have my SSH KEY on
>>>> the
>>>>> target machine.
>>>>
>>>> I am OK with this.
>>>>
>>>>> As I don't know your network, maybe what you could do is: Via
>>>> RasperBerry,
>>>>> forward the SSH to the target machine, I will pass-through via your
>>>>> RasperBerry where you can control the access for the rest of your LAN.
>>>>>
>>>>> Another approach could be, two different subnets and a firewall. Or as
>>>> you
>>>>> said, VLANS.
>>>>
>>>> I will be doing VLANS, which have yet to be set up.
>>>>
>>>> The target system will have ZFS pools can be configured for different
>>>> tests (i.e. raidz2 vs raidz3).
>>>> This will involve gpart etc because the drives & pools will need to be
>>>> 'wiped' between different test
>>>> runs.
>>>>
>>>> I seem to recall someone suggesting PXE boot and configuring the system
>>>> remotely.  Does anyone
>>>> recall that?  That aspect of the discussion was not recorded:
>>>> https://wiki.freebsd.org/201510DevSummit/Performance
>>>>
>>>>
>>> Bapt@ mentioned that, this is the way how we are doing in another
>> project.
>>> But in my point of view, it is not a must for our case!
>>>
>>> The PXE wold be good if we try to test different of OS flavors, or build
>>> different images.
>>
>> OK.  The only thing holding us back is:
>>
>> - adding the air filters to the case
>> - moving to the new switch with the new VLANs
>>
>> It's now a matter of time.
>>
>> —
>> Dan Langille
>> http://langille.org/
>>
>>
>>
>> _______________________________________________
>> [hidden email] mailing list
>> https://lists.freebsd.org/mailman/listinfo/freebsd-performance
>> To unsubscribe, send any mail to "
>> [hidden email]"
>>
>
>
> Dan,
>
> I don't know if you have already cracked the serial access nut, but every
> supermicro motherboard I have owned has supported serial over lan. I
> personally have never used it, I have just noticed it from time to time.
> Here is a good resource on setting it up
> http://serverfault.com/questions/574351/serial-over-lan-on-freebsd-10-0-with-supermicro-x9-scm-f
> (note: it's for an X9 motherboard but it should work for your X10, see the
> comment at the very end)

If we go that way, I will keep that in mind, thank you.

> Also, FreeBSD, ZFS and performance are three of my most favorite words. Is
> there any way I can participate in this adventure?

Yes, there will be, I'm sure.

This comes to mind immediately: running the tests on your own server will validate/invalidate our results.

For now, we have no results, because it's all waiting on me and my network at home.  It's a time issue.
This weekend is pretty much filled with other stuff.


Dan Langille
http://langille.org/





_______________________________________________
[hidden email] mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-performance
To unsubscribe, send any mail to "[hidden email]"
Reply | Threaded
Open this post in threaded view
|

Re: Measuring ZFS configuration differences

Dan Langille
In reply to this post by Dan Langille

> On Nov 18, 2015, at 4:12 PM, Dan Langille <[hidden email]> wrote:
>
>
>> On Nov 17, 2015, at 9:31 PM, Marcelo Araujo <[hidden email]> wrote:
>>
>> 2015-11-18 10:23 GMT+08:00 Dan Langille <[hidden email]>:
>>
>>>
>>>> On Nov 17, 2015, at 8:46 PM, Marcelo Araujo <[hidden email]>
>>> wrote:
>>>>
>>>> 2015-11-18 3:14 GMT+08:00 Dan Langille <[hidden email]>:
>>>>
>>>>>
>>>>> On Nov 12, 2015, at 1:30 AM, Marcelo Araujo <[hidden email]>
>>>>> wrote:
>>>>>
>>>>>
>>>>>
>>>>> 2015-11-12 6:34 GMT+08:00 Dan Langille <[hidden email]>:
>>>>>
>>>>>> On Oct 12, 2015, at 1:00 PM, Dan Langille <[hidden email]> wrote:
>>>>>>>
>>>>>>> Following up on the discussions during EuroBSDCon 2015 (Stockholm)
>>>>>> during the FreeBSD Developer
>>>>>>> Summit regarding various ZFS configuration settings, I write to start
>>>>>> our implementation phase now that some
>>>>>>> usual suspects have joined the list.
>>>>>>>
>>>>>>> re https://wiki.freebsd.org/201510DevSummit/Performance
>>>>>>>
>>>>>>> I think the first order of business is granting access rights to the
>>>>>> server (varm) in question:
>>>>>>>
>>>>>>> http://dan.langille.org/2015/07/19/varm/
>>>>>>>
>>>>>>> During the workshop, mention was made of serial access.  I can arrange
>>>>>> that.
>>>>>>>
>>>>>>> The server has IPMI, however, my first thought:
>>>>>>>
>>>>>>> 1 - connect a USB-serial cable to varm & link that to another server
>>> in
>>>>>> my rack.
>>>>>>
>>>>>> Marcelo: At EuroBSDCon, was it you who mentioned a particular
>>>>>> configuration for the test machine which made
>>>>>> it easy to configure and run tests?  Was it PXE booting or something?
>>>>>>
>>>>>>> 2 - create a jail in that server and give it access to that serial
>>>>>> connection
>>>>>>> 3 - redirect incoming port XYZ to that jail via a public-key-only ssh
>>>>>> connection
>>>>>>> 4 - give people access
>>>>>>>
>>>>>>> Any suggestions?
>>>>>>
>>>>>> —
>>>>>> Dan Langille
>>>>>> http://langille.org/
>>>>>>
>>>>>>
>>>>> Hello Dan,
>>>>>
>>>>> Yes, was me :)
>>>>>
>>>>> I mention about zopkio test framework.
>>>>> I gave a presentation last weekend at PyCon Hong Kong about it.
>>>>>
>>>>> Here is my slides:
>>>>>
>>> http://www.slideshare.net/araujobsd/functional-and-scale-performance-tests-using-zopkio
>>>>>
>>>>> The good of Zopkio is, we can write tests at once and run it as much as
>>> we
>>>>> want in different machines. Also Zopkio depends of Naarad, that can
>>> parse a
>>>>> CSV file and create metrics and SLA over those metrics, plot graphs and
>>> so
>>>>> on. Pretty nice tool!!!
>>>>>
>>>>> I'm wondering if we could start to test something and maybe show it at
>>>>> AsiaBSDCon and BSDCon(Canada) next year? What do you think?
>>>>> What I need right now would be a list of tests that we want to perform
>>> as
>>>>> well as what parameters we would like to take as metrics to compare.\
>>>>>
>>>>>
>>>>> For tests, we can start with this list:
>>>>> https://github.com/dlangille/zfs_benchmarks/issues
>>>>>
>>>>> We can start as soon as I figure out how to provide access to the
>>>>> testers.  See above re serial connection.
>>>>>
>>>>> I want to provide access, but I want to keep access restricted to only
>>>>> this box and not to the rest of my home LAN.  I plan to do this via a
>>>>> VLAN.
>>>>>
>>>>> I could fire up a Rasperberry Pi and allow ssh into that.  Will that be
>>>>> enough
>>>>> power for what you need to do?
>>>>>
>>>>>
>>>> First of all, thanks to share the tests cases.
>>>>
>>>> If I use zopkio, the best would be access SSH direct to the target
>>> machine
>>>> where I need to run the tests. For zopkio, I need to have my SSH KEY on
>>> the
>>>> target machine.
>>>
>>> I am OK with this.
>>>
>>>> As I don't know your network, maybe what you could do is: Via
>>> RasperBerry,
>>>> forward the SSH to the target machine, I will pass-through via your
>>>> RasperBerry where you can control the access for the rest of your LAN.
>>>>
>>>> Another approach could be, two different subnets and a firewall. Or as
>>> you
>>>> said, VLANS.
>>>
>>> I will be doing VLANS, which have yet to be set up.
>>>
>>> The target system will have ZFS pools can be configured for different
>>> tests (i.e. raidz2 vs raidz3).
>>> This will involve gpart etc because the drives & pools will need to be
>>> 'wiped' between different test
>>> runs.
>>>
>>> I seem to recall someone suggesting PXE boot and configuring the system
>>> remotely.  Does anyone
>>> recall that?  That aspect of the discussion was not recorded:
>>> https://wiki.freebsd.org/201510DevSummit/Performance
>>>
>>>
>> Bapt@ mentioned that, this is the way how we are doing in another project.
>> But in my point of view, it is not a must for our case!
>>
>> The PXE wold be good if we try to test different of OS flavors, or build
>> different images.
>
> OK.  The only thing holding us back is:
>
> - adding the air filters to the case
> - moving to the new switch with the new VLANs
>
> It's now a matter of time.

Done.

The server is ready for access.

Please send me your ssh-key off-list and I will create a login for you.

We do have some data on the system:

$ zpool list
NAME            SIZE  ALLOC   FREE  EXPANDSZ   FRAG    CAP  DEDUP  HEALTH  ALTROOT
benchmarking   21.8T  6.44T  15.3T         -    11%    29%  1.00x  ONLINE  -
music            30T  5.58T  24.4T         -     9%    18%  1.00x  ONLINE  -
random_mirror  2.72T   372K  2.72T         -     0%     0%  1.00x  ONLINE  -
zroot           220G  1.79G   218G         -     0%     0%  1.00x  ONLINE  -

The music pool has about 1.4TB of music files for testing via copy.


Dan Langille
http://langille.org/





_______________________________________________
[hidden email] mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-performance
To unsubscribe, send any mail to "[hidden email]"
Reply | Threaded
Open this post in threaded view
|

Re: Measuring ZFS configuration differences

Marcelo Araujo-7
Hey Dan!

2015-11-25 2:13 GMT+08:00 Dan Langille <[hidden email]>:

>
>
> Done.
>
> The server is ready for access.
>

Cool!!! Very good news, thank you to work on it and provide resources!


>
> Please send me your ssh-key off-list and I will create a login for you.
>

I will do very soon!


>
> We do have some data on the system:
>
> $ zpool list
> NAME            SIZE  ALLOC   FREE  EXPANDSZ   FRAG    CAP  DEDUP  HEALTH
> ALTROOT
> benchmarking   21.8T  6.44T  15.3T         -    11%    29%  1.00x  ONLINE
> -
> music            30T  5.58T  24.4T         -     9%    18%  1.00x  ONLINE
> -
> random_mirror  2.72T   372K  2.72T         -     0%     0%  1.00x  ONLINE
> -
> zroot           220G  1.79G   218G         -     0%     0%  1.00x  ONLINE
> -
>
> The music pool has about 1.4TB of music files for testing via copy.
>

By the name, I guess the benchmark(ing) is the one we will use for our
tests.


Best,
--

--
Marcelo Araujo            (__)[hidden email]
\\\'',)http://www.FreeBSD.org <http://www.freebsd.org/>   \/  \ ^
Power To Server.         .\. /_)
_______________________________________________
[hidden email] mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-performance
To unsubscribe, send any mail to "[hidden email]"
Reply | Threaded
Open this post in threaded view
|

Re: Measuring ZFS configuration differences

Dan Langille

> On Nov 24, 2015, at 9:00 PM, Marcelo Araujo <[hidden email]> wrote:
>
> Hey Dan!
>
> 2015-11-25 2:13 GMT+08:00 Dan Langille <[hidden email]>:
>
>>
>>
>> Done.
>>
>> The server is ready for access.
>>
>
> Cool!!! Very good news, thank you to work on it and provide resources!
>
>
>>
>> Please send me your ssh-key off-list and I will create a login for you.
>>
>
> I will do very soon!
>
>
>>
>> We do have some data on the system:
>>
>> $ zpool list
>> NAME            SIZE  ALLOC   FREE  EXPANDSZ   FRAG    CAP  DEDUP  HEALTH
>> ALTROOT
>> benchmarking   21.8T  6.44T  15.3T         -    11%    29%  1.00x  ONLINE
>> -
>> music            30T  5.58T  24.4T         -     9%    18%  1.00x  ONLINE
>> -
>> random_mirror  2.72T   372K  2.72T         -     0%     0%  1.00x  ONLINE
>> -
>> zroot           220G  1.79G   218G         -     0%     0%  1.00x  ONLINE
>> -
>>
>> The music pool has about 1.4TB of music files for testing via copy.
>>
>
> By the name, I guess the benchmark(ing) is the one we will use for our
> tests.
I want to retain music intact.  I think everything else can be reconfigured as required.

We will be doing a lot of reconfiguration: raidz1-3, 1-13 drives in an array, etc.

For now, let's make sure you can login, ensure your root access works, etc.


Dan Langille
http://langille.org/






signature.asc (990 bytes) Download Attachment