Posts by Keith E. Laidig

21) Message boards : Rosetta@home Science : RE: Rosetta@home hardware/fireware failure - Sept. 8/07 (Message 45827)
Posted 9 Sep 2007 by Profile Keith E. Laidig
Post:
It is truly unfortunate to have such an event occur to the Rosetta@Home project. I, and the Rosetta@Home global user community are very proud to be a part of this project from the beginning, and we will be here to assist this project for as long as it is running.

We would like to give a piece of our mind to the vendor of the hardware though... but I think David and the Rosetta@Home team will do a good job giving our piece of mind to them anyway.

David, let us know when everything is up and running.
The Rosetta world community is here to back you up 100%.

Regards,

John Koulouris,(Esq.)


I have to admit that it is tempting to give out the email address of the OEM vice presidents to the R@H community to allow the user base to express their opinion directly to those concerned... -KEL
22) Message boards : Cafe Rosetta : Last one to post here wins! #2 (Message 45778)
Posted 9 Sep 2007 by Profile Keith E. Laidig
Post:
Any idea when my hungry crunchers will have some work? :)


We need the magic of DK's expertise to get the work flowing out normally... He should be on this in the next few few hours....

Sorry for the delay. -KEL
23) Message boards : Number crunching : Welcome Back! (Message 45754)
Posted 9 Sep 2007 by Profile Keith E. Laidig
Post:
You People at the Project have been doing one heck of a GREAT JOB! We will see you when we see you.

Mike F,


I appreciate your patience but we're embarrassed.... I plan to pass along my discomfort to a couple of OEM vice presidents next week!

Post if anything doesn't work as you expect. -KEL
24) Message boards : Cafe Rosetta : Last one to post here wins! #2 (Message 45752)
Posted 9 Sep 2007 by Profile Keith E. Laidig
Post:
me too... let me tell you! -KEL
25) Message boards : Cafe Rosetta : Word link 10 (Message 44667)
Posted 3 Aug 2007 by Profile Keith E. Laidig
Post:
Bill Haley and The Comets
Comet Halley
comet - the song
Colonel Bogey March
fore!
golf
Los Angeles Country Club
playboy mansion
Playboy Clubs
Playboy bunny
easter bunny
Easter Eggs
Milka
Nestle
Swiss Miss (Hot Chocolate)
Swiss Chocolate
Swiss birthday (Swiss National Day: 1st August; Switzerland was founded in the Swiss Alpes)
come on guys, keep it in one thread, how else do you "word link" properly?
swiss weather = swiss alps

















26) Message boards : Rosetta@home Science : Does RAH have any ties to Fred Hutchinson Cancer research group? (Message 44332)
Posted 26 Jul 2007 by Profile Keith E. Laidig
Post:
It's funny you should ask - in addition to the existing collaborations w/ Barry Stoddard and Ray Monnat at FHCRC see the discussion here, one of the developers of Rosetta just started a new job at 'the Hutch'. Dr Phil Bradley took up his new position in the Computational Biology Group of the Public Health Sciences Division at the beginning of July.

As to the details of the various projects, I'll leave that to the scientists to outline...

do you guys get any information from them or send anything to them from what you have learned on any cancer projects to date or in the future?

27) Message boards : Number crunching : Unusual connection attempts FROM srv3.bakerlab.org (Message 44114)
Posted 23 Jul 2007 by Profile Keith E. Laidig
Post:
I'm giving SRV3 a close look now. This appears to be a caching issue (the server trying to address requests for webpages some long period of time after the request was made)... more when I finish up.

Thanks for the information on the server's behavior... -KEL

Update

Well, SRV3 is back to it's usual/proper behavior (I took this downtime/opportunity to update the OS as well). The cause of the apparent troubles was a tremendous backlog of webpages requests that SRV3 was unable to manage in a timely manner. As a result, the machine appeared - in specific instances - to be sending out unrequested packets...

The origin of the trouble is our preparation to upgrade the back-end fileserver to R@H. Our good RAID/SCSI fileserver is just barely able to keep up with all of your efforts. David B has decided to invest in replacing this fileserver with a fiber SAN solution and we've been trying to prepare for the change over by synchronizing the existing data onto the SAN (more details once we're done). This synch'ing of the data is hard on the system and can lead to large loads on the various components - as we saw last night.

I apologize to the RAH user base for this (unanticipated) problem. We try very hard not to cause any problem on the user end...

We'll try to get the move to the SAN done very soon!

-KEL

Had to reboot my router this morning. Seems I'm getting flooded with connection attempts from srv3.bakerlab.org. Here's a short sample from the first five minutes after rebooting the router.

Mon Jul 23 04:33:40 2007 Blocked access attempt from 140.142.20.111:80 to TCP port 57281
Mon Jul 23 04:33:40 2007 Blocked access attempt from 140.142.20.111:80 to TCP port 57283
Mon Jul 23 04:33:49 2007 Blocked access attempt from 140.142.20.111:80 to TCP port 57293
Mon Jul 23 04:33:52 2007 Blocked access attempt from 140.142.20.111:80 to TCP port 57293
Mon Jul 23 04:33:58 2007 Blocked access attempt from 140.142.20.111:80 to TCP port 57293
Mon Jul 23 04:33:58 2007 Blocked access attempt from 140.142.20.111:80 to TCP port 57297
Mon Jul 23 04:34:01 2007 Blocked access attempt from 140.142.20.111:80 to TCP port 57297
Mon Jul 23 04:34:07 2007 Blocked access attempt from 140.142.20.111:80 to TCP port 57297
Mon Jul 23 04:34:10 2007 Blocked access attempt from 140.142.20.111:80 to TCP port 57293
Mon Jul 23 04:34:19 2007 Blocked access attempt from 140.142.20.111:80 to TCP port 57297
Mon Jul 23 04:34:28 2007 Blocked access attempt from 140.142.20.111:80 to TCP port 57281
Mon Jul 23 04:34:28 2007 Blocked access attempt from 140.142.20.111:80 to TCP port 57283
Mon Jul 23 04:34:34 2007 Blocked access attempt from 140.142.20.111:80 to TCP port 57293
Mon Jul 23 04:34:43 2007 Blocked access attempt from 140.142.20.111:80 to TCP port 57297
Mon Jul 23 04:35:22 2007 Blocked access attempt from 140.142.20.111:80 to TCP port 57293
Mon Jul 23 04:35:31 2007 Blocked access attempt from 140.142.20.111:80 to TCP port 57297
Mon Jul 23 04:35:52 2007 Blocked access attempt from 140.142.20.111:80 to TCP port 57313
Mon Jul 23 04:35:55 2007 Blocked access attempt from 140.142.20.111:80 to TCP port 57313
Mon Jul 23 04:36:00 2007 Blocked access attempt from 140.142.20.111:80 to TCP port 57317
Mon Jul 23 04:36:01 2007 Blocked access attempt from 140.142.20.111:80 to TCP port 57313
Mon Jul 23 04:36:03 2007 Blocked access attempt from 140.142.20.111:80 to TCP port 57317
Mon Jul 23 04:36:09 2007 Blocked access attempt from 140.142.20.111:80 to TCP port 57317
Mon Jul 23 04:36:13 2007 Blocked access attempt from 140.142.20.111:80 to TCP port 57313
Mon Jul 23 04:36:21 2007 Blocked access attempt from 140.142.20.111:80 to TCP port 57317
Mon Jul 23 04:36:37 2007 Blocked access attempt from 140.142.20.111:80 to TCP port 57313
Mon Jul 23 04:36:45 2007 Blocked access attempt from 140.142.20.111:80 to TCP port 57317
Mon Jul 23 04:37:25 2007 Blocked access attempt from 140.142.20.111:80 to TCP port 57313
Mon Jul 23 04:37:33 2007 Blocked access attempt from 140.142.20.111:80 to TCP port 57317
Mon Jul 23 04:38:21 2007 Blocked access attempt from 140.142.20.111:80 to TCP port 57339
Mon Jul 23 04:38:21 2007 Blocked access attempt from 140.142.20.111:80 to TCP port 57341

28) Message boards : Cafe Rosetta : R@H Bumper Stickers, what do you think? (Message 42037)
Posted 10 Jun 2007 by Profile Keith E. Laidig
Post:

bumper sticker on profile...............


...on a beautiful Mustang! You are a true supporter! I'm impressed.... -KEL
29) Message boards : Cafe Rosetta : R@H Bumper Stickers, what do you think? (Message 41924)
Posted 7 Jun 2007 by Profile Keith E. Laidig
Post:
You might want to try out CafePress.com for this. It's easy to create what you like...

-KEL
30) Message boards : Rosetta@home Science : How can we bring more users to the Rosetta project? (Message 41145)
Posted 19 May 2007 by Profile Keith E. Laidig
Post:
Once things settle down I'll update the technical details of everything so that those of you who might be interested can see how we do it here in the Baker Lab - for BOINC and more generally at our datacenters in the Seattle area.

Keith,

I understand you are busy. So, please don't over tax yourself to get us this information. However, I must say, it sounds interesting. I sit anxiously waiting for your post. ;-)

Thanks,

Joel


Joel,

I'll get to this soon. Sorry for the delay...


-KEL
31) Message boards : Rosetta@home Science : How can we bring more users to the Rosetta project? (Message 40711)
Posted 11 May 2007 by Profile Keith E. Laidig
Post:

I agree w/ you. We, as a group, don't have a great deal of time to figure out how to best do this. Can you provide guidance to a project that has done one or more of these things well so that we can avoid re-inventing the wheel? I'd appreciate it!

Thanks,

-KEL

I wasn't disputing what you said in general, but as a veteran of many boards across many subjects, having a Wiki, FAQ, embedded encyclopedia with all the worlds knowledge, etc., will not stop people asking their question on the boards. It is therefore I only quoted that part of the message.

People are lazy, they don't want to look for information, they want the information to come to them.

32) Message boards : Rosetta@home Science : How can we bring more users to the Rosetta project? (Message 40440)
Posted 6 May 2007 by Profile Keith E. Laidig
Post:
For newer projects and their more experimental codes we use our local resources. We have invested a significant amount of money over the past five years into the resources here in Seattle.

Hence Keith's RAC of 120k+!


My climb up the BOINC lists are going to be short lived. We're just finishing up a burn in of our newest super-cluster and will open it up to the scientists this coming week. Once they take over my score will drop like a stone.

I hate not having our computers working so I use BOINC to soak up every idle CPU cycle here as well. No idle CPU cycles around here!

Once things settle down I'll update the technical details of everything so that those of you who might be interested can see how we do it here in the Baker Lab - for BOINC and more generally at our datacenters in the Seattle area.
33) Message boards : Rosetta@home Science : How can we bring more users to the Rosetta project? (Message 40390)
Posted 6 May 2007 by Profile Keith E. Laidig
Post:
But then the shocker: "projects are not currently running on BOINC because we don't yet have an efficient queuing system"


I think David was saying that we don't have ALL of our projects running on BOINC. What is running now IS valuable research!

We've tended to use BOINC to tackle our more mature lines of research with more mature code - so as to avoid mis-using the contributions of the BOINC community. The last thing we want to do is waste your contributions on shoddy code or poorly/incorrectly implemented algorithms!

For newer projects and their more experimental codes we use our local resources. We have invested a significant amount of money over the past five years into the resources here in Seattle. This is what we apply to code development, algorithmic testing, new ideas, and exotic projects. Only after we are confident that the resources would be well spent do we submit it to BOINC.

The specific issue is one of data management: How do we create a work flow system that is robust enough to handle the wide variety of different research problems as well as handle the stupendous amount of data that is returned from BOINC? This is what we're working on today...
34) Message boards : Number crunching : Failed download (Message 39440)
Posted 15 Apr 2007 by Profile Keith E. Laidig
Post:

fixed. -KEL
35) Message boards : Number crunching : Failed download (Message 39177)
Posted 8 Apr 2007 by Profile Keith E. Laidig
Post:
Problem: The network interfaces on srv3 went offline again.

Solution: network interfaces cycled off and on again.

Long-term issues: We're asking the UW to increase the bandwidth to the download servers and we've been tuning the NIC drivers' TSO features to address the NIC's stability issues.

We'll get this problem cracked sooner or later... -KEL
36) Message boards : Number crunching : Failed download (Message 38839)
Posted 2 Apr 2007 by Profile Keith E. Laidig
Post:
Thanks for the heads up - restored the interfaces. It looks like we're getting unwanted attention that is causing problems. I've instituted new technology that should help. it's always something... -KEL
37) Message boards : Number crunching : Failed download (Message 38702)
Posted 29 Mar 2007 by Profile Keith E. Laidig
Post:

Folks,

The webservers are back online - but, as you might imagine, there something of a backlog of download requests. The servers are throwing out data as fast as they can. It might take a while, but we should catch up here in a few hours...


-KEL
38) Message boards : Number crunching : Failed download (Message 38638)
Posted 29 Mar 2007 by Profile Keith E. Laidig
Post:

We're looking into this.. -KEL
39) Message boards : Number crunching : Delays on downlaod - file server struggling - or is it me? (Message 37744)
Posted 13 Mar 2007 by Profile Keith E. Laidig
Post:
two of the download servers had similar hiccups - kernel faults - resulting in two of the four IP addresses being unresponsive to download requests. Sorry 'bout that, we're working on making things more robust.... -KEL
40) Message boards : Rosetta@home Science : Rosetta@home store (Message 29727)
Posted 21 Oct 2006 by Profile Keith E. Laidig
Post:
This sort of thing has been brought up before.... the last time of which I am aware was here:

http://boinc.bakerlab.org/rosetta/forum_thread.php?id=2350#28622

and in the associated discussion... It all depends (it seems to me) upon

a) the amount of interest from the R@H user-base, and
b) how the proceeds could/would/should be treated by the UW and/or HHMI....

-KEL


Previous 20 · Next 20



©2024 University of Washington
https://www.bakerlab.org