Science and Technology news... :: View topic

Brian · Guest

G>day all,

I will be going overseas for 3 1/2 weeks & I want to keep my 2 pc>s
crunching SETI while I am gone.

I am using SetiGate & I figured that I would download enough work units for
4 weeks, doing my last send/receive just before leaving for the airport & do
a send receive as soon as I get home.

Would this a good or bad idea?

I have thought about the SetiGate auto connect, but I don>t know how to set
up a once per week dialup for my modem.
My other idea is to drive 50km (30 miles) & have friends look after my pc>s
& SETI for me.

All this, just to keep crunching ;-)

Brian
3848 units

FalconFly · Guest

[quote]I will be going overseas for 3 1/2 weeks & I want to keep my 2 pc>s
crunching SETI while I am gone.
[/quote]
If you have sufficient Fire, Electrical and realtime Hardware Monitoring protection (or
someone to look after your machines at least once a day), not a bad Idea.
It really depends on how much you trust your machines, and how well they are prepared for
a worst case scenario.

Just imagine a Power Supply going bad and delivering faulty voltages to the Motherbard, a
few days after you leave :p
(in ~182000 operating hours so far, this happened to me twice with AT based machines, and
once I exchanged a somewhat odd/disturbingly smelling ATX Power Supply ahead of time).

In both cases, the failure was catastrophic and destroyed the System.
On one of the occasions, I woke up by a beeping Bios Alert in the middle of the night and
noted already intense smell of smoldering PCB. Within maybe 2-3 Minutes, the Voltage
Regulators of the AT Motherboard were on their way to melt through the PCB :p
On the other occasion, the System was Shutdown in less than a minute after the Bios Alarm
went off.
Still, the CPU & its Cooler reached in excess of 150 Degrees, and the Motherboard
Capacitors were about to blow their caps.

So generally, I would never leave an AT based System running unattended for more than 1-2
days maximum, since there>s nothing (except external devices or sensors) that can shut
them down in case of trouble.

A Fire hazard is usually very remote with computers, but I never put this worst-case
catastrophy completely out of consideration, since the risk does exist.

[quote]I am using SetiGate & I figured that I would download enough work units for
4 weeks, doing my last send/receive just before leaving for the airport & do
a send receive as soon as I get home.

Would this a good or bad idea?
[/quote]
4 weeks cache is way too big.
All Units returned will be long processed by other Users, leaving the Results useful for
your Stats counter only.

I>d recommend you configure your Server for Auto-Dialin a minimum of once every 48hrs.
I used to do this when I was away from home for more than a week, and it worked very nice.

[quote]I have thought about the SetiGate auto connect, but I don>t know how to set
up a once per week dialup for my modem.
My other idea is to drive 50km (30 miles) & have friends look after my pc>s
& SETI for me.
[/quote]
Depends on OS your running.
If you>re using Windows, simply set it in the Browser and/or Dialup settings.
There, you can also specify after how many minutes of inactivity your Server logs off
again.

But of course, you should test this setup before you leave ;)

Greetings
FalconFly

Webmaster
http://www.falconfly.de
3dfx Archive
====================
Forum : http://www.falconfly-central.de

Buck Rogers · Guest

[quote]
Depends on OS your running.
If you>re using Windows, simply set it in the Browser and/or Dialup
settings.
There, you can also specify after how many minutes of inactivity your
Server logs off
again.

But of course, you should test this setup before you leave ;)

Greetings
FalconFly

Webmaster
http://www.falconfly.de
3dfx Archive
====================
Forum : http://www.falconfly-central.de

[/quote]
Hey Falconfly, did you manage to install the Linux on your Flagship dual
Athlon? :)
Or did NT save the day? *grin*

Buck

FalconFly · Guest

[quote]Hey Falconfly, did you manage to install the Linux on your Flagship dual
Athlon? :)
Or did NT save the day? *grin*

Buck
[/quote]
Hehe, after lots of tweaking, I managed to have it run at about 95% of the competing
Win2000 Dual box ;)

Yesterday, I reverted from V3.08 to V3.03 on basically all my Boxes, and the two Dualies
are now about equal it seems, with the Linux Box having a slight edge (numbers have to
stabilize to make a judgement).

Still a pity the Linux Clients are so much slower, but at least it is now the fastest Box
in the house ;)
(althought 'cheating' with its 400MHz bonus *g*)

That performance I can live with, after all a free Operating System means to take a few
compromises.
So my evil plans to convert from Windows to Linux are still on track :D

Greetings
FalconFly

Webmaster
http://www.falconfly.de
3dfx Archive
====================
Forum : http://www.falconfly-central.de

Martin · Guest

FalconFly wrote:
[...]
[quote]Still a pity the Linux Clients are so much slower, but at least it is
now the fastest Box in the house ;) (althought 'cheating' with its
400MHz bonus *g*)

That performance I can live with, after all a free Operating System
means to take a few compromises. So my evil plans to convert from
Windows to Linux are still on track :D
[/quote]

I>m now running the SetiQueue - SetiSpy - Win-NT-CLI-3.03 combination
under Wine on Mandrake 9.1 Linux. Seem to be getting marginally better
times than when running Windoze although there may be a few distortions
due to the reobservation units that have gone through. Might also be due
to more accurate time keeping/logging.

Very nice to be able to dual boot and have the Seti stuff pick up from
where it left off unperturbed.

But then again, not needed to go back to Windoze for some time now...
May well squish those partitions down somewhat...

(Also, s@h are listing new tapes being split after a 3 week hiatus.
Still getting old WUs through though at the moment.)

Keep crunchin'
Martin

--
----------
- Martin -
- 53N 1W -
----------

FalconFly · Guest

[quote](Also, s@h are listing new tapes being split after a 3 week hiatus.
Still getting old WUs through though at the moment.)

Keep crunchin'
Martin
[/quote]
Off-Topic, but I noted the same.
The last batches of WU>s must have circulated for something like 2 weeks, smelled like at
least 15x-25x redundancy :p

Anyway, lucky enough they found someone to split some other Tapes after all.

Webmaster
http://www.falconfly.de
3dfx Archive
====================
Forum : http://www.falconfly-central.de

FalconFly · Guest

[quote]What is so evil about converting from Window$ to Linux? Did it years
ago. Last Win version used in our house was 3.11. I have a HD with
win 98 on it but it hasn>t seen power for over a year.
[/quote]
Well, that was rather some form of insider joke ;)

[quote]Still running on P90 at home which is slow but in some cases it is
faster than my Windows 98 machine at work. Getting a new computer
this fall, finally. I have a new computer at work that came with XP
pro and I have been using RH 8.0 since day one in a dual boot mode.
SetiQueue runs 5 to 6 WU>s a day at work using 3.08.
[/quote]
Lack of time was basically what kept me from switching to Linux earlier.
While I can assemble, setup and fully optimize a Windows box in a matter of hours, the
same still takes me about a week for a Linux Box (internet searches, troubleshooting,
internet searches again to find a 'correct' HowTo among the many incomplete/incorrect or
simply unusable ones, fighting insane Text editors like 'vi' etc etc... ;) )

Haven>t managed to make them run exactly with the same features yet, but until the end of
the year, the existing Installations should be finetuned to what I need.

But anyway, I>m getting there...
It just eats tons of time and manually setting everything up, unlike Windows.

The only thing I>m not happy at all about is the User friendlyness (quick reconfiguring,
installing/modifying Software), which I think has barely approached Windows 3.1 levels now
:(
(e.g. overall GUI gives only the most basic features, but most of the powerful
applications/services completely lack their own simple GUI>s [which is a big mistake
IMHO], and require extensive manual fiddling on Terminal level :p )

Anyway, for setting them up once, and basically not touching them afterwards, they>re
still very good.

Webmaster
http://www.falconfly.de
3dfx Archive
====================
Forum : http://www.falconfly-central.de

Antonis Markouizos · Guest

G>morning...

It would be a really nice Idea. I did it last week (one week only vacations,
but it did work prerfectly)

Just set also the pc to auto -reboot (also set "Halt on" no errors) in case
of a power fail, and auto logon (as a user) if you have such issues.
U can dissable additional software i.e. firewalls and antivirus, if u are on
windows xp, to shutdown internet explorer from task manager and whatever is
used by you but not by the pc),
U can also disconect cd-roms and additional hard drives, add a small cooler
to the hard drive seti will run on and let the case open for better cooling
(a nice excuse for the harddrives and cdroms is at laaast to reduse the
bill, the heat,the decay from the wear, the and accelerate a liiitle bit
the system)

Autoconnect when u are away, with a dialup is a bad idea though you may have
problems on connection
(line drooped when sending or recieving), and many other things. since the
cost is per hour you may face a pretty ungly suprise when you 'll be back :D

Sending a friend for 50km just to look a pc would be unkind, and difficult
(to explain how, and specially WHY) even he/she is willing to
(after all thats my my opinion)

Download workunits for 4-5 weeks and let it go.
(dialup? 56k? you should let it download for the whole night, and as many as
it gets. i.e. set it to 500,
and on the morning set it to "finish" one or two "after" those it got down)

Since you are already using setigate, you wont have to do anythink special,
just download the wu>s

have a nice time,

A.

Ï "Brian" <tramsmash@yahoo.com> Ýãñáøå óôï ìÞíõìá
news:3f24f010_1@news.iprimus.com.au...
[quote]G>day all,

I will be going overseas for 3 1/2 weeks & I want to keep my 2 pc>s
crunching SETI while I am gone.

I am using SetiGate & I figured that I would download enough work units
for
4 weeks, doing my last send/receive just before leaving for the airport &
do
a send receive as soon as I get home.

Would this a good or bad idea?

I have thought about the SetiGate auto connect, but I don>t know how to
set
up a once per week dialup for my modem.
My other idea is to drive 50km (30 miles) & have friends look after my
pc>s
& SETI for me.

All this, just to keep crunching ;-)

Brian
3848 units

[/quote]

Martin · Guest

Antonis Markouizos wrote:
[...]
[quote]Download workunits for 4-5 weeks and let it go.
(dialup? 56k? you should let it download for the whole night, and as many as
it gets. i.e. set it to 500,
and on the morning set it to "finish" one or two "after" those it got down)
[/quote]
_BIG_ QUEUES ARE BAD FOR SETI@HOME!!!

If your queue delays you returning the result for a WU from the TIME OF
DOWNLOADING THAT WU for too long, that result is near useless. (And very
expensive electricity to just increment your WU count.)

Also, the s@h server must keep a pool of WUs that matches the pool of
WUs temporarily lost in everyone>s queues, to then later check if enough
results have been returned before clearing a particular WU from the
pool. Big user queues increase the workload on the s@h database and server.

As to how big a queue is reasonable?...

Various vague comments have been made in the past. There seems to be a
concensus that you should have you queue store no more than 1 week>s
supply of WUs.

My opinion is that you should store no more than two day>s supply of WUs
to allow for the rare occasion that s@h>s servers are offline over a
weekend.

The recent reobservations WUs appear to have caused a bit of a glitch in
WU distribution... (Users preferentially queuing reobservation WUs into
big queues? (;-))

Keep crunchin'
(with a reasonable queue size for your WU crunching rate)
Martin

--
----------
- Martin -
- 53N 1W -
----------

Martin G. Diehl · Guest

Martin wrote:
[quote]
Antonis Markouizos wrote:
[...]
Download workunits for 4-5 weeks and let it go.
(dialup? 56k? you should let it download for the whole
night, and as many as it gets. i.e. set it to 500, and
on the morning set it to "finish" one or two "after"
those it got down)

_BIG_ QUEUES ARE BAD FOR SETI@HOME!!!
[/quote]
unsupportable conclusion

[quote]If your queue delays you returning the result for a WU
from the TIME OF DOWNLOADING THAT WU for too long, that
result is near useless.
[/quote]
Not necessarily true ... that "late" WU result could be
the tie-breaker.

[quote](And very expensive electricity to just increment your
WU count.)

Also, the s@h server must keep a pool of WUs that matches
the pool of WUs temporarily lost in everyone>s queues, to
then later check if enough results have been returned
before clearing a particular WU from the pool.
[/quote]
False. The S@H WU server does not wait for any results
before clearing a WU -- it sends each WU to several clients
initially. This has been known for about 2 years.

[quote]Big user queues increase the workload on the s@h database
and server.
[/quote]
False ... no evidence for this conclusion.

[quote]As to how big a queue is reasonable?...
[/quote]
Nobody knows.

[quote]Keep crunchin'
(with a reasonable queue size for your WU crunching rate)
Martin

--
----------
- Martin -
- 53N 1W -
----------
[/quote]
--
Martin G. Diehl

Reality -- That which remains after you stop thinking
about it.

All replies and comments accepted and considered.
Adoption of your suggestions are at my sole discretion.
Award criteria are unpublished and are considered to
be a trade secret. As such, awards for your responses
cannot be guaranteed. Incoming flames may be stored
for use during severe winters or may be circulated for
deep analysis, peer review, and/or literary criticism.

FalconFly · Guest

There used to be an official Posting of the SAH Administration several month ago :

Ideal Cache Size : 24-48hrs max.
(Results get fully accounted into Database, since most Results return within that
Timeframe)

Max. desireable Cache : ~7 days
(some of the Results are obsolete already, but since some never return at all, still
limited useful for science. Most of those old Results also serve "Database Integrity"
checks)

Greater than 1 week (e.g. upto 1 month) :
Almost fully redundant.
No scientific use, almost all Results have the 'verified' Tag already in the Database. Too
old even for Integrity checks, and increasing the stats counter only ...

This is what I recall, and the average Return rate will sure have slightly dropped again
within the last months when the statement was made.

The idea of big, long running local Caches was fully supported a long time ago, when
average Return rates were alot lower (and the gross computing power to SAH much smaller as
well); and the Server often was overloaded, dropping lots of connections on a
regular/normal basis.

But these times are definitely over.

I therefor limited my Caches to max. ~24hrs long ago, since one has to add the actual
computing time, and the delay to the next Cache Flush onto it as well (e.g. a machine
working 8hrs will return a Result it after a total of 32hrs minimum, no delays included)
With the Congent Uplink mostly working flawless, I think that>s a good technique.

So IMHO, 'getting them back home' should happen ideally a rough 36hrs after reception.

Greetings
FalconFly

Webmaster
http://www.falconfly.de
3dfx Archive
====================
Forum : http://www.falconfly-central.de

Martin · Guest

Martin G. Diehl wrote:
[quote]Martin wrote:
[...]

_BIG_ QUEUES ARE BAD FOR SETI@HOME!!!

unsupportable conclusion
[/quote]
On the contrary, that statement is very supportable. Think further...
FalconFly>s very good answer should give you a few clues... (Or just do
a google for some good past discussions.)

[quote]If your queue delays you returning the result for a WU
from the TIME OF DOWNLOADING THAT WU for too long, that
result is near useless.

Not necessarily true ... that "late" WU result could be
the tie-breaker.
[/quote]
Very true... However, how late is too late?

[quote](And very expensive electricity to just increment your
WU count.)

Also, the s@h server must keep a pool of WUs that matches
the pool of WUs temporarily lost in everyone>s queues, to
then later check if enough results have been returned
before clearing a particular WU from the pool.

False. The S@H WU server does not wait for any results
before clearing a WU -- it sends each WU to several clients
initially. This has been known for about 2 years.
[/quote]
Your "False" is wrong. Your following statement however is true...

[quote]Big user queues increase the workload on the s@h database
and server.

False ... no evidence for this conclusion.
[/quote]
Try again... Think how s@h works (for the SCIENCE -- big hint (;-))

[quote]As to how big a queue is reasonable?...

Nobody knows.
[/quote]
An easy calculation given the server specs vs running stats. We on the
outside can make some good educated guesses.

....
Warning... Answer spoiler warning...

....

....

....

....

....

....!

If s@h was just counting returned WUs and then throwing them away to
just award one point per WU, then your apparent assumptions and comments
would be completely correct.

Reality is -- Those WUs which remain after you have cross-checked for at
least 3 returns or more. (And the database bashing and WU disk space
required to support that for 600 000 active users multiplied by the size
of their queues.)

Onwards with a more efficient (small queue delayed) crunch...

Regards,
Martin

--
----------
- Martin -
- 53N 1W -
----------

Martin · Guest

Martin wrote:
[quote]
(And very expensive electricity to just increment your WU count.)
[/quote]

Also note that if your WU result is returned later than four other
results for that WU, your result is of little additional scientific
value. (You still get credited with a WU result being returned ofcourse.)

Return a result after the WU pool has been churned, and that WU result
is just thrown away. (You still get credited with a WU result being
returned ofcourse.)

My guess is that at present, if you can get WU results returned within
about a week of downloading, your efforts should be helping science.

Otherwise, there are other worthwile projects out there that can better
accept long return delays for their results.

Otherwise, save generating some waste heat.

....And then again, this project can be a good education!

[quote]Onwards with a more efficient (small queue delayed) crunch...

Regards, Martin
[/quote]

--
----------
- Martin -
- 53N 1W -
----------

Brian · Guest

After reading all the comments, I have learned (& a little more confused)
about caching workunits. I think that from now on I will send/receive
workunits 2 times per week.

I have decided against letting my pc>s to run unsupervised for the 3 1/2
weeks that I will be away. I want to come home to 2 working pc>s.
Instead I will go for the 'friends to look after' option. One of them is now
looking forward to having another pc so he can LAN game at home & my father
will look after my other pc.

Thanks to all who replied

Brian
3888 units

Antonis Markouizos · Guest

Ok, guess, I didnt Know It.

Since Seti Give a 4 week "limit" before deactivating you, I thought that may
be something similar for the work units.
It seems that things have changed.....

Well, Shouldnt be developed a "system of "tracking" the WUs?
I mean, since a member is continiusly returning some results per week, if he
downloads 50-60 WU (1-2 weeks "job" for a 2 pc system), isnt really
possible that he will return 'em? The WU have I think something to identify
them. If I download some of them ie from one "number" to one other, and I>m
returning results from that "collection" Its really possible that I>ll
return 'em all
(using something like SetiGate)
Ok, I guess that here are people returing 60-600wu per day, and for them its
like kiding saying for my 2-8 results per day) but I believe the most of the
people dont belong to these Seti User "Class". I also know that a system
like this will also require some more cpu power, that Seti team may find
difficult to find, But It would set the numper current numper of the active
users really more efficient.Since there are people that may find difficult
to sit down and watch out the pc when it I>ll finish to sent the results,
(I>m working, and I>m using the seti on the two pc>s home. Since the
connection is a dialup with a really limited bandwith and a per hour cost, I
have to download the WUs the time I really dont use connection nor the
telephone, but I>m home and on the Pc after 10p.m., and I dont have to wake
up next morning before the sun does.)

well is just some thoughts. any answers are appreciated

A.

(as u may have noticed, I>m not a proffessor on English Philology....)

? "Martin" <ml_news@ddnospamddml1dd.co.uk.dd> ?????? ??? ??????
news:5r6Wa.1750$yl6.710@newsfep4-winn.server.ntli.net...
[quote]Antonis Markouizos wrote:
[...]
Download workunits for 4-5 weeks and let it go.
(dialup? 56k? you should let it download for the whole night, and as
many as
it gets. i.e. set it to 500,
and on the morning set it to "finish" one or two "after" those it got
down)

_BIG_ QUEUES ARE BAD FOR SETI@HOME!!!

If your queue delays you returning the result for a WU from the TIME OF
DOWNLOADING THAT WU for too long, that result is near useless. (And very
expensive electricity to just increment your WU count.)

Also, the s@h server must keep a pool of WUs that matches the pool of
WUs temporarily lost in everyone>s queues, to then later check if enough
results have been returned before clearing a particular WU from the
pool. Big user queues increase the workload on the s@h database and
server.

As to how big a queue is reasonable?...

Various vague comments have been made in the past. There seems to be a
concensus that you should have you queue store no more than 1 week>s
supply of WUs.

My opinion is that you should store no more than two day>s supply of WUs
to allow for the rare occasion that s@h>s servers are offline over a
weekend.

The recent reobservations WUs appear to have caused a bit of a glitch in
WU distribution... (Users preferentially queuing reobservation WUs into
big queues? (;-))

Keep crunchin'
(with a reasonable queue size for your WU crunching rate)
Martin

--
----------
- Martin -
- 53N 1W -
----------
[/quote]


www.GetXFactor.com Leading Technology, Science, Agriculture News and information	Part of the Identityscape.com network... getxfactor.com jmoodmusic.com smartbusinesschoices.com mintdepot.com lowfaresalways.com evangelicalview.com shoppingpodder.com soproudlywehail.com webnews.ws currenthumor.com

Science and Technology news... Forum Index -> Astrology - SETI Forum	Goto page 1, 2 Next
Page 1 of 2