Take a Pebble (Dec 18 2007)


log in

Advanced search

Message boards : Technical News : Take a Pebble (Dec 18 2007)

1 · 2 · Next
Author Message
Profile Matt Lebofsky
Volunteer moderator
Project administrator
Project developer
Project scientist
Avatar
Send message
Joined: 1 Mar 99
Posts: 1389
Credit: 74,079
RAC: 0
United States
Message 692650 - Posted: 18 Dec 2007, 23:24:47 UTC

Our Tuesday outage ran a little long this week because we're no longer dumping to the super fast Snap Appliance as we converted that space into more workunit storage. Instead we're currently writing to the internal disk space on thumper, which is vast but much slower for some reason. This situation will evolve, so nothing really to worry about.

We also made the database change to fix the cryptic bug noted in this thread. Pretty much just adding a new row to the middle of the application table so it was in sync with the data structs in the code. And yep, after that it was behaving normally, even without our "force" to set values to where they should be regardless of what was erroneously culled from the database. So we're calling this fixed.

I also got the new server "bane" on line as a third redundant public web server. Perhaps you noticed a speedup? Perhaps you noticed some unexpected garbage, broken links, or weird php behavior? Let me know via this thread if you see anything obviously (and suddenly) wrong with the web site. Over the coming days we will retire the current web servers kosh and penguin. Bane is a system with two Intel quad-core 2.66GHz CPUs and 4GB RAM in 1U of rack space. Alone it is more powerful than kosh and penguin combined, which together account for about 6U of rack space.

- Matt

____________
-- BOINC/SETI@home network/web/science/development person
-- "Any idiot can have a good idea. What is hard is to do it." - Jeanne-Claude

Profile eaglescouter
Send message
Joined: 28 Dec 02
Posts: 162
Credit: 42,012,553
RAC: 0
United States
Message 692654 - Posted: 18 Dec 2007, 23:46:49 UTC - in response to Message 692650.

Retrieval of the "my computers" page is quite sluggish (but it does work)
____________
It's not too many computers, it's a lack of circuit breakers for this room. But we can fix it :)

Anthony Apfelbeck
Avatar
Send message
Joined: 26 May 06
Posts: 72
Credit: 2,865,363
RAC: 3,602
United States
Message 692661 - Posted: 19 Dec 2007, 0:38:49 UTC - in response to Message 692654.

Retrieval of the "my computers" page is quite sluggish (but it does work)


Same for me. The technical news and account pages are taking a very long time to load.
____________
Join the Fire Service Team

Brian Silvers
Send message
Joined: 11 Jun 99
Posts: 1681
Credit: 492,052
RAC: 0
United States
Message 692662 - Posted: 19 Dec 2007, 0:40:13 UTC - in response to Message 692650.

Our Tuesday outage ran a little long this week because we're no longer dumping to the super fast Snap Appliance as we converted that space into more workunit storage. Instead we're currently writing to the internal disk space on thumper, which is vast but much slower for some reason. This situation will evolve, so nothing really to worry about.


With the "pebble" in the thread title, I don't want you to think that I'm "stoning you", but have you passed on the information about the deadline situation that Richard Haselgrove and I brought up a week or so ago? I would think that if you reduce the amount of time given (the deadline) for tasks down to something less than what they are today, yet still generous, you would keep slow / less-allocated hosts from building up as large of a work cache so that faster / more-allocated hosts that returned their results quicker would likely not have their returned results lingering in storage so much, thus decreasing the overall storage needs.

As for the web server performance, right now is not a good time to evaluate that... I'll give a yea/nay later...
____________

PhonAcq
Send message
Joined: 14 Apr 01
Posts: 1622
Credit: 22,369,488
RAC: 6,177
United States
Message 692664 - Posted: 19 Dec 2007, 0:52:32 UTC

Feedback:
1) public web server isn't just slow it is adiabatic (at the time of this post)
2) 8 Intel cores with only 4GB of RAM? You must not be running Windows!

OzzFan
Volunteer tester
Avatar
Send message
Joined: 9 Apr 02
Posts: 13625
Credit: 31,034,617
RAC: 20,705
United States
Message 692695 - Posted: 19 Dec 2007, 4:27:59 UTC - in response to Message 692664.

Feedback:
1) public web server isn't just slow it is adiabatic (at the time of this post)
2) 8 Intel cores with only 4GB of RAM? You must not be running Windows!


Windows is too costly for a project running on fumes for funding. They're switching over to Linux on most of their servers.
____________

Brian Silvers
Send message
Joined: 11 Jun 99
Posts: 1681
Credit: 492,052
RAC: 0
United States
Message 692722 - Posted: 19 Dec 2007, 7:20:27 UTC
Last modified: 19 Dec 2007, 7:39:57 UTC

Well, it has been 6.5 hours since I posted saying I'd comment on the web server performance later, and the pages are still slow to load...

[edit]
15 minutes after I tried to post the original message, it was still trying to reload the thread. I had been off looking into my 401(k) and not watching the forum.

So, "slow" is not even close to being the correct term for how abysmal the forum pages are performing...

[/edit]
____________

Brian Silvers
Send message
Joined: 11 Jun 99
Posts: 1681
Credit: 492,052
RAC: 0
United States
Message 692734 - Posted: 19 Dec 2007, 8:13:22 UTC - in response to Message 692730.

I dunno......I just got home from working an 11 hour shift, and at the time of this posting, the webpage response in all the forums is verrrry snappy!


Well, I dunno either... It is not my connection... My connection is advertised as 1.5Mbit... (slowest of the plans available), and here's speedtests to San Jose:



Los Angeles:



and San Francisco:

Speedy
Volunteer tester
Avatar
Send message
Joined: 26 Jun 04
Posts: 678
Credit: 5,926,770
RAC: 3,988
New Zealand
Message 692742 - Posted: 19 Dec 2007, 8:36:03 UTC - in response to Message 692650.
Last modified: 19 Dec 2007, 8:45:36 UTC

Our Tuesday outage ran a little long this week because we're no longer dumping to the super fast Snap Appliance as we converted that space into more workunit storage.

All togeather how mutch storage is there for wu's now?
As I write this the fourm is rather sloooow
cheer's

Speedy

Profile Pooh Bear 27
Volunteer tester
Avatar
Send message
Joined: 14 Jul 03
Posts: 3221
Credit: 2,640,394
RAC: 1,530
United States
Message 692768 - Posted: 19 Dec 2007, 10:51:31 UTC - in response to Message 692662.
Last modified: 19 Dec 2007, 10:51:52 UTC

With the "pebble" in the thread title, I don't want you to think that I'm "stoning you", but have you passed on the information about the deadline situation that Richard Haselgrove and I brought up a week or so ago? I would think that if you reduce the amount of time given (the deadline) for tasks down to something less than what they are today, yet still generous, you would keep slow / less-allocated hosts from building up as large of a work cache so that faster / more-allocated hosts that returned their results quicker would likely not have their returned results lingering in storage so much, thus decreasing the overall storage needs.

This has been discussed in this forum by Matt. Matt is not the person who controls the length. He has passed on the concern, but he personally can not do anything about it.

Read over some of the past couple weeks of messages in this forum.
____________

PhonAcq
Send message
Joined: 14 Apr 01
Posts: 1622
Credit: 22,369,488
RAC: 6,177
United States
Message 692801 - Posted: 19 Dec 2007, 14:01:08 UTC - in response to Message 692695.

Feedback:
1) public web server isn't just slow it is adiabatic (at the time of this post)
2) 8 Intel cores with only 4GB of RAM? You must not be running Windows!


Windows is too costly for a project running on fumes for funding. They're switching over to Linux on most of their servers.


Please accept my apologies for my lame attempt at geek humor.

OzzFan
Volunteer tester
Avatar
Send message
Joined: 9 Apr 02
Posts: 13625
Credit: 31,034,617
RAC: 20,705
United States
Message 692808 - Posted: 19 Dec 2007, 15:16:25 UTC - in response to Message 692801.

Feedback:
1) public web server isn't just slow it is adiabatic (at the time of this post)
2) 8 Intel cores with only 4GB of RAM? You must not be running Windows!


Windows is too costly for a project running on fumes for funding. They're switching over to Linux on most of their servers.


Please accept my apologies for my lame attempt at geek humor.


No apologies needed. Sorry it went over my head! LOL :-)
____________

Profile arkaynProject donor
Volunteer tester
Avatar
Send message
Joined: 14 May 99
Posts: 3689
Credit: 48,728,280
RAC: 6,436
United States
Message 692821 - Posted: 19 Dec 2007, 16:05:59 UTC

I see Matt is listening to some Emerson Lake & Palmer now.
____________

Ben, Shell & Bobby
Send message
Joined: 8 Aug 99
Posts: 2
Credit: 107,340
RAC: 0
Australia
Message 692828 - Posted: 19 Dec 2007, 17:01:08 UTC

Hi Matt, I have very recently returned to Seti after a long break, and I'm not really sure of how the credit system works yet, but since the outage, when I look at my stats (although the total credit seems right) in the your results I seem to have lost most of the work units done on the 17th. And the number of results returned for each computer seems to have gone down a little?


____________

Profile Dr. C.E.T.I.
Avatar
Send message
Joined: 29 Feb 00
Posts: 15993
Credit: 690,597
RAC: 0
United States
Message 692850 - Posted: 19 Dec 2007, 18:50:32 UTC



< Thanks for the Post Matt . . . so, as 'arkayn' mentioned previously, ELP's playin' right below ;)

ELP performs @ 1974 California Jam . . . "Take A Pebble"

Date: Sat, 25 Dec 2004 14:28:21 -0200
From: Fabio
Subject: e/emerson_lake_and_palmer/take_a_pebble.tab


'Take a Pebble' - Emerson Lake And Palmer
Tabbed by Fabio Gaspar

This tab is for the quiet part in the middle. I think it was just Greg
Lake improvisation in D, but i tabbed it. Is just a little part and is
not written in proper timing.


e-------------------5----------------------------------
b----------3--5/7/-------------------------------------
g----2--4----------------------------------------------
d-0----------------------------------------------------
a------------------------------------------------------
e------------------------------------------------------

e-------------------5--7---7--5------------------------
b---------3--5/7/----------------7--5--3---------------
g---2--4-----------------------------------------------
d-0----------------------------------------------------
a------------------------------------------------------
e------------------------------------------------------

e------------------5--7/-10--10\\7\\5-(5)/7--------------
b---------3--5/7/--------------------------------------
g---2--4-----------------------------------------------
d-0----------------------------------------------------
a------------------------------------------------------
e------------------------------------------------------


Then you can do all you want in D (i guess)

____________
BOINC Wiki . . .

Science Status Page . . .

Profile Matt Lebofsky
Volunteer moderator
Project administrator
Project developer
Project scientist
Avatar
Send message
Joined: 1 Mar 99
Posts: 1389
Credit: 74,079
RAC: 0
United States
Message 692880 - Posted: 19 Dec 2007, 21:40:28 UTC - in response to Message 692828.

Hi Matt, I have very recently returned to Seti after a long break, and I'm not really sure of how the credit system works yet, but since the outage, when I look at my stats (although the total credit seems right) in the your results I seem to have lost most of the work units done on the 17th.


Welcome back! In order to keep the BOINC database nice and trim (for faster I/O) we only keep 24 hours of completed results around before purging them, which is why they disappeared from your list. The credit, science, etc. of course remains - we're just removing the excess database records.

- Matt
____________
-- BOINC/SETI@home network/web/science/development person
-- "Any idiot can have a good idea. What is hard is to do it." - Jeanne-Claude

zoom314Project donor
Avatar
Send message
Joined: 30 Nov 03
Posts: 46517
Credit: 36,857,421
RAC: 5,033
United States
Message 692944 - Posted: 20 Dec 2007, 1:34:29 UTC - in response to Message 692749.
Last modified: 20 Dec 2007, 1:36:27 UTC

Oh, and thank you for the link.....
Very nice bandwidth test site!!!

Yes It is isn't It Doc. ;) Yeah the website seems to be doing pretty good right now.

____________
My Facebook, War Commander, 2015

1 · 2 · Next

Message boards : Technical News : Take a Pebble (Dec 18 2007)

Copyright © 2014 University of California