Posts by ric

11) Message boards : Észrevételek, tapasztalatok : Átállás 11 dimenzióra / We are about to change to 11 dimension (Message 348)
Posted 4467 days ago by ric
>Sorry but we have no patience. We need our daily quota of WUs.... ;-)

Thierry you have my fully understanding.

On social and mental level.

Be a gentlemen and support other projects, for example the nearly unknown project of SETI@home.
Really, a world rank of 8 is a extrem bad number, this should be worked out...

Be strong and let's wait for monday.

BTW tks for keeping us informed..




12) Message boards : Észrevételek, tapasztalatok : Átállás 11 dimenzióra / We are about to change to 11 dimension (Message 344)
Posted 4467 days ago by ric
When the boinc clients acts well, most of the time it does..,
it's up the client to get as much as it's possibilities offers.

The client is (AFAK) not asking for a number of work, the client is asking for
a certain number of hours of work.

In this case, it doesn't matter, if there is a p III or a faster host running.

In theory, the maximum Queue length/size (connect every..)is 4.0 Days for this project. So again in theory, there is no longer wait for more than those 4 day to wait until every result is returned.

From my experiences, sometimes the estimated CPU time is much higher than the
effective elapsed time. I guess, even a work log of 4 Days is done in real 2 Days.

As long as the resulting work is not needed to be cross verified to get a
coherance/zero toleranz of 0.00000000000000001 (like LHC...) it can be considered that redundant work is a waste of cpu cycles.

A project can progress faster without the multiple send out the same work.

As usual, the coin has several sides and therefore several opinions are valid;)

Anyway, there are several (boinc-based)projects arrount to keep the clients busy until SKTAKI will lounch/generate the next work stage.

The shares are an effective way to "manage" the workflow.

Let the dev people get the time they need, may I remember, it's only "a couple of days ago" since we had the honnor to crunch the very first work of this project..

waitting patiently
13) Message boards : Észrevételek, tapasztalatok : Problem with WU upload (Message 334)
Posted 4469 days ago by ric
Probaly my bad.

Was thinking, the river called Bosporus would pass trough Budapest.

No it's the River called Donau.

While digging more the net, something new to learn, in erlier days,
the towns buda and pest(and Óbuda ?) have been town for themself, later
they merged it to the town to "Budapest".

One of the oldest available subways is also serving Budapest.

Just to be mentioned, sorry for touristic view, Budapest is an old Town, over 50'000 years old, it has a remarkable history, turkish bath Zoo and beerfactory from the 16th centaury. Beatifull parks.

Váci utca /shopping mile?

and something like as Várhegy

And old town with character...


(go BWM-Sauber)
14) Message boards : Észrevételek, tapasztalatok : Problem with WU upload (Message 329)
Posted 4470 days ago by ric
Hungarian Dialects

Alfold, West Danube, Danube-Tisza, King's Pass Hungarian, Northeast Hungarian, Northwest Hungarian, Székely, West Hungarian. Closest to Vogul (Mansi) of Russia. Speakers of Standard Hungarian have difficulty understanding Oberwart dialect

Population 10,298,820 in Hungary (1995). Population total all countries: 13,611,600.
Region Also spoken in Australia, Austria, Canada, Israel, Romania, Serbia and Montenegro, Slovakia, Slovenia, Ukraine, USA.
Alternate names Magyar


<a href="http://www.ethnologue.com/show_language.asp?
code=hun">http://www.ethnologue.com/show_language.asp?code=hun</a>

for one year, at school, we had a hungarian teacher for geography/politology, he was a nice person and told us much stories from his escape on early day

Please don't forget <b>Zsa Zsa Gabor</b>, one of the most known Hungarian woman..

And Budapest with the bosporus
15) Message boards : Észrevételek, tapasztalatok : Átállás 11 dimenzióra / We are about to change to 11 dimension (Message 314)
Posted 4471 days ago by ric
finished the work.

goodluck for next stage
16) Message boards : Észrevételek, tapasztalatok : Számolási hiba (computing error) (Message 268)
Posted 4474 days ago by ric
is the understanding right, you have a couple of PCs who are not connected to the internet, but you have an USB based network "plug" and with this you are
walking from host to host to give a temporary internet access??

Basically an interesting idea, but perhaps a little time demanding..

The OS will survive this torture, but what's an level application?

The OS is ready for this handling, what will be brocken is the "logical session" from the clients application(s) to the project servers.

With a defined dial in connection, the client (should) know, when a physical connection is established or not and can handle it.

At your place, the USB device there is "just no more available", the behave could differ.

Sure you did it, prior pluging off the USB network connector, you did the "disable BOINC network access" to the client and only when getting the message "Suspending network activity - user request" then pluging off hte USB device.

might it could be 2 other things:

try a more recent driver for the USB based network adapter/USB hub, perhaps a driver bug..

Perhaps in the properties of the driver, there is a setting covering better the powering on/off of the device and it's (logical) session management

goodluck!!!

17) Message boards : Észrevételek, tapasztalatok : Everytime only 1 WU (Message 266)
Posted 4474 days ago by ric
&gt;Consider it a gift.

LOL!

as long as it's a free gift...

Yesterday, unfortunately an old harddisk crashed and died for ever.

Very happy to had a Queue (connect every..) of only 1.9 Days.

So I "only" wasted about 54 work units, what I'm very sorry for.

This is not like other projects with a lot of anounced and not anounced downtimes, even a connect every 0.4 Days can be used without running out of work.

Crossing all fingers, but this project is running really stable in the last 2 Weeks or so.
This is a good omen;)
18) Message boards : Észrevételek, tapasztalatok : Számolási hiba (computing error) (Message 265)
Posted 4474 days ago by ric
0xc0000043=STATUS_SHARING_VIOLATION

This error message can be taken as a direction where to dig.

"take care" to use not to much over clocked hardware...

a ping or traceroute on networklevel could show how long a network packet
needs to "walk" over the network.

When a network interface, a router (software and hardware), a proxyserver or something in this area is "slowing down" the handling, the origin client could suffer.


goodluck!

--&gt; same issue when using a more recent boinc client, like 4.44, 4.45??
19) Message boards : Észrevételek, tapasztalatok : Sum of Rec.Av.Credits weird (Message 260)
Posted 4474 days ago by ric
PoorBoy is right with his opinion, RAC means nothing, it's just a momentary
element.

Just finished a test, wanted to keep the network connections stopped until until 500 WUs have been done.

Unfortunately 2 PCs have been running out so I had to re-enable the network connections.

Right now the RAC was taking a jump, due about 470 finished WUs reported back.

I guess the only thing what can be taken from the RAC is, when you crunch "much" and the RAC is not moving.
Then this could mean, the calculation itself is "not ok".

This happened to me at LHC, one host was getting only 1/3 of RAC which an other host of the same speed class made.

Reducing the OC speed on this low RAC based host was also reducing/eleminating the "cpu error" and finaly the results could be done "ok".
Finaly the RAC of this corrected host was climbing.

As mentioned, it was at LHC, I don't know the behaviour of this project in term of "quality" of work.

happy crunching


20) Message boards : Észrevételek, tapasztalatok : Credits ? (Message 194)
Posted 4482 days ago by ric
Adam,

please let me add to your statement,

when in case of exception, the work is returned with "0" cputime/credits, the RAC will not climb, perhaps even falling.

Your are right, the RAC is a kind of <b>actual</b>speed.

More interesting could be, can the resulting RAC of this project be compared to the RAC of other project?.

Not quite sure, IMHO it looks like, with the same hardware used for crunching, the possible RAC here is deeper as for example at SETI.




Next 10 posts

Home | My Account | Message Boards


Copyright © 2017 SZTAKI Desktop Grid