Mailing List Archive

David Gerard wrote:
> On 10/10/06, Brad Patrick <bradp.wmf@gmail.com> wrote:
>
>
>> If we have +/- 250 servers running now
>> *and we double in traffic every four months
>> *and the costs per server (for round number purposes, including racks,
>> power, etc.) are $4000
>> *we are talking about *starting* with a $1M computer order being
>> fulfilled before March 1.
>> If we add to that the next 500 computers, at the same rate, we are
>> talking about a batch of $2M beyond that. /me's head spins.
>>
>
>
> Heh. What's the upfront and service contract on an IBM minicomputer -
> make that two clustered in different cities - running as many copies
> of Linux/390 as we need?
>
> (I wonder if IBM would sponsor that, just to say they had.)
>
>
> - d.
>
>

IBM don't sell their mainframes as low-cost compute farms, and that's
because mainframes simply aren't _that_ powerful relative to their
price; what they are useful for is consolidating lots and lots of
relatively lightly-loaded services into one _highly_ reliable and
maintainable system. That reliability and maintainability is what you
pay for when you buy a mainframe; and if you really need it, you need to
get a mainframe. (There are, for instance, stories of mainframes being
rebuilt like George Washington's axe until none of the original machine
was left, whilst still in service, without stopping any programs from
running or taking any filesystems down: they were simply shifted from
processor to processor, and disk to disk, with the system still running.)

For example, a fully-loaded 54-CPU top-of-the range IBM z9-109
"enterprise-class" mainframe is claimed to be able to do "a billion
transactions per day". Now, if we assume that each of those transactions
is equivalent to serving a Wikipedia hit, that's about 11000 hits per
second, about 60% of the load of 18000 hits per second currently served
by the several hundred CPUs in the Wikipedia cluster. However, if the
price list below is anything to go by, it'll set you back about $22M for
the privilege, which is considerably more than the entire Wikimedia
Foundation budget to date.

http://www-306.ibm.com/software/swnews/swnews.nsf/n/lsan6euhj9?OpenDocument&Site=swzseries

http://www.tech-news.com/publib/pl2094.html

However, if you wanted a high-price/performance-ratio compute farm from
IBM, I'm sure they could oblige in other ways: for example, they
currently hold the first, second, and third places in the TOP500
supercomputer listing with their closely-coupled cluster systems, so I
would imagine they could probably quote you for a web server farm.

You could always ask them for a quote: as a high-prestige project,
there's always a possibility they might be inclined to do a good deal on
it for the PR benefit... but I'm willing to bet they'll try to sell you
a cluster system rather than a mainframe.

-- Neil


_______________________________________________
foundation-l mailing list
foundation-l@wikimedia.org
http://mail.wikipedia.org/mailman/listinfo/foundation-l

Erik Moeller wrote:
>
> Tim - when are the static dumps going to be ready?
>
It's out of disk space. The images are using over 600GB already, not
including image tarballs which will require that much again. Also I'm
out space for text, srv31's 230GB is full. The project really needs
either a dedicated 2TB or some software changes to reduce disk usage.

So in answer to your question, I was intending on waiting until the new
image servers arrive. Then I could try to grab one to finish the present
dump. Then I would do the software changes and run the subsequent dump
on a lower-specced machine. I don't know when the image servers will
arrive, but we're all hoping it will be soon.

-- Tim Starling
_______________________________________________
foundation-l mailing list
foundation-l@wikimedia.org
http://mail.wikipedia.org/mailman/listinfo/foundation-l

We need a fund drive focused on robustness and reliability.

Discussion about what new projects to promote is important, but we still
lack medium-term and long-term plans for making the most basic maintenance
of the projects robust and self-sustaining. There are serious puzzles and
issues related to technical scalability, with theoretical solutions, which
would benefit from more attention. And part of robustness is robustness
under budget cuts.

Two+ years ago, I proposed five goals that I hoped the foundation would
address. One of these was developing an infrastructure that could support
a thousand million hits a day with 99.9% uptime, in a self-sustaining
fashion that only required $50k/yr and 1 full-time employee for upkeep.
http://meta.wikimedia.org/wiki/Three-year_plan

We are now close to a thousand million hits a day, with improving uptime;
but have made little progress in slowing, not to mention reversing, the
growth of annual upkeep (yahoo's kindness last spring has not yet become
the first of many).

I firmly believe that the foundation should be focusing on reducing the
annual cost and complexity of running core operations, and that the
Wikimedia server cluster needs an endowment. This becomes more true
when people within the community start to feel Wikipedia is 'old hat'
and a solved problem, and become fixed on attacking new problems before
wrapping up that old one.

Basic needs which an endowment could support:

* buying the most basic infrastructure: bandwidth, machines and other
hardware (that which cannot be / has not yet been donated)
* a project to work on the scalability and reliability of the WP
infrastructure (including the capacity to accept more distributed donations
of hardware and bandwidth)
* a project to pursue bandwidth and hardware/maintenance contributions.
With proper attention, considering our support in the tech community, we
should be able to eliminate these expenses -- and safeguard their
continued coverage for the next five or ten years.


On 10/9/06, Anthere <Anthere9@yahoo.com> wrote:

>As an individual (not board member), I would like to ask if it were
>possible that the spin around the fundraising, focuses a bit on the
>other projects, perhaps the notion of virtual library, with the
>wikicommons and wikisource. We have the recent report of the german
>digitization to show up as an example as how we could push things
>forward in realm of digital libraries.

Creating a virtual library, by improving commons and wikisource, is great.
Focusing on the smaller projects, likewise. But to make 'pushing things
forward' the spin/goal of a fundraiser, when our current dreams & projects
are not yet secure, may be like putting the cart before the horse.


Brad writes:
> If we have +/- 250 servers running now
> *and we double in traffic every four months
> *and the costs per server (for round number purposes, including racks,
> power, etc.) are $4000
> *we are talking about *starting* with a $1M computer order being
> fulfilled before March 1.
>
> If we add to that the next 500 computers, at the same rate, we are
> talking about a batch of $2M beyond that. /me's head spins.

These are the most important and largest expenses at the moment, and growth
will not slow over the next few years unless we collectively fail. But
rather than worrying about whether growth will level out, we should find
ways to dispose of them.


On 10/10/06, Anthere <Anthere9@yahoo.com> wrote:

> I am not favorable to the concept of making a joint fundraising with
> another organisation. As for OLPC, they were already part of our last
> press release in august...

I've never heard of a 'joint fundraising' with multiple groups; neither
do I see why WM should want such a thing. Wikipedia in particular is a
primal force of our networked society - sustaining its success is as
lofty a fundraising goal as could be hoped for. This is why we have
individuals making monthly $1000 contributions, even when there is no
fund drive. The projects do not need to primp and showcase 'new' ideas
or initiatives.

What they need is clear articulation of their raison d'etre [WP is good
about this; other projects less so], a good description of what is needed
to sustain and propagate current success, and an explanation of how funds
will help provide what is needed.

--SJ

_______________________________________________
foundation-l mailing list
foundation-l@wikimedia.org
http://mail.wikipedia.org/mailman/listinfo/foundation-l

On 10/11/06, SJ Klein <meta.sj@gmail.com> wrote:
> I firmly believe that the foundation should be focusing on reducing the
> annual cost and complexity of running core operations, and that the
> Wikimedia server cluster needs an endowment. This becomes more true
> when people within the community start to feel Wikipedia is 'old hat'
> and a solved problem, and become fixed on attacking new problems before
> wrapping up that old one.

Responding briefly, more later.

1) If we agree that we need a very large amount of money for hardware,
and if we do agree that we want to raise that money from the community
at large, then we should of course communicate that, using clear facts
and figures.

2) Even so, it remains true that:
- There are now several Foundation employees or equivalents, including
a CEO, who all get salaries. We need to communicate what our staff is
doing, besides taking care of the hosting and administration.
- Money going to chapters will not necessarily have to go into
Foundation-level investments, but could also support local projects
and salaries.
- Hardware is not something people generally get very excited about.
Jimmy's personal appeal in the last fundraiser was very effective.
However, if we repeat the same kind of appeal, we also need to have a
little more to show for it this time. Hence my plea for reports and
information, which we should collect in any case, regardless of the
specific strategy we follow.

3) It may not necessarily be the best thing to raise all or even any
of the money for hardware from the community at large. An endowment,
as you say, grants, or partnerships may carry us a long way. $1M is a
lot of money for us -- but for some who might support us, it is very
little indeed. The question, of course, is -- can we raise the
necessary funds from other sources as quickly as we require them?

--
Peace & Love,
Erik

Member, Wikimedia Foundation Board of Trustees

DISCLAIMER: Unless otherwise stated, all views or opinions expressed
in this message are solely my own and do not represent an official
position of the Wikimedia Foundation or its Board of Trustees.
_______________________________________________
foundation-l mailing list
foundation-l@wikimedia.org
http://mail.wikipedia.org/mailman/listinfo/foundation-l

On 10/10/06, daniwo59@aol.com <daniwo59@aol.com> wrote:
> Will the money go to chapters or to the Foundation? Can the chapters
> transfer money to the foundation if needed?

This is a big open question. As I understand it, the current arrangement is:
1) Most money goes directly to WMF.
2) Chapters act as local fundraising agents, e.g. in Germany, where
many people don't have credit cards, a bank transfer directly to the
German chapter can be made. Some individuals might also choose to
support a chapter instead of the WMF for other reasons.
3) Some or all of the money raised through 2) (is supposed to) go back
to the WMF.

Is this correct? What is the precise status on 3)? My view is that
chapters should retain a substantial amount of donations for local
projects and infrastructure. How do we decide that amount? What is the
best strategy for transferring funds to the WMF -- could royalties for
the trademark/logo work?

> How will funds donated be processed? Who will process them? How will we
> recognize donors, both by what is required by law and by what is simply a nice
> way to recognize them? This is an enormous amount of work.

These are specifics that are best worked out with the FundCom at a
dedicated meeting that should take place very soon.
--
Peace & Love,
Erik

Member, Wikimedia Foundation Board of Trustees

DISCLAIMER: Unless otherwise stated, all views or opinions expressed
in this message are solely my own and do not represent an official
position of the Wikimedia Foundation or its Board of Trustees.
_______________________________________________
foundation-l mailing list
foundation-l@wikimedia.org
http://mail.wikipedia.org/mailman/listinfo/foundation-l

Erik Moeller wrote:

>On 10/10/06, daniwo59@aol.com <daniwo59@aol.com> wrote:
>
>
>>Will the money go to chapters or to the Foundation? Can the chapters
>>transfer money to the foundation if needed?
>>
>>
>This is a big open question. As I understand it, the current arrangement is:
>1) Most money goes directly to WMF.
>2) Chapters act as local fundraising agents, e.g. in Germany, where
>many people don't have credit cards, a bank transfer directly to the
>German chapter can be made. Some individuals might also choose to
>support a chapter instead of the WMF for other reasons.
>3) Some or all of the money raised through 2) (is supposed to) go back
>to the WMF.
>
>Is this correct? What is the precise status on 3)? My view is that
>chapters should retain a substantial amount of donations for local
>projects and infrastructure. How do we decide that amount? What is the
>best strategy for transferring funds to the WMF -- could royalties for
>the trademark/logo work?
>
National laws need to be considered. Under Canadian law maintaining
charitable status depends on spending a nearly all the money for
charitable purposes in the country. If we ever get to that stage
domestic projects should be considered. In some cases receiving
royalties as business income could be considered as taxable income.

Ec

_______________________________________________
foundation-l mailing list
foundation-l@wikimedia.org
http://mail.wikipedia.org/mailman/listinfo/foundation-l