Discussion Forums

Community Meeting 2012-09-25

1 reply [Last post]
jrseti
jrseti's picture
Offline
Joined: 2010-07-22
Posts: 250

 

IRC chat on Tuesday! Join us September 25 2012, 11am California time, 18:00GMT, for our weekly 30 minute setiQuest IRC chat at irc://irc.freenode.net/#setiquest
 
Or in a browser: http://webchat.freenode.net/?channels=setiquest
Today's topic can be:

  1. The new air conditioning is installed and working at the ATA. I'll provide an update.
  2. I am building another SonATA distro. I'll discuss. Still having problems!
  3. Discuss the future of this IRC chat. See http://setiquest.org/forum/topic/future-irc-meetings

Any other ideas are appreciated.

-jrseti

 

-jrseti

sigblips
sigblips's picture
Offline
Joined: 2010-04-20
Posts: 732
[11:00am] jrseti_:

[11:00am] jrseti_: Hello!
[11:00am] jrseti_: See http://setiquest.org/forum/topic/community-meeting-2012-09-25
[11:00am] gliese581c: Title: Community Meeting 2012-09-25 | setiQuest (at setiquest.org)
[11:01am] sigblips: Hi there.
[11:01am] welterde: sigblips: yeah.. run linux.. and I meant more for your cuda work(the frontend nodes each have a tesla in them)
[11:01am] jrseti_: I assume Lou from SETILive is not on the chat?
[11:02am] sigblips: Nope, but Lou did a brief IRC test several hours ago.
[11:02am] welterde: sigblips: and I think the current setup is that each display is connected to one node
[11:02am] jrseti_: What's this about CUDA?
[11:03am] sigblips: welterde: OK. So using it for a remote X11 display testing? Not for the large display purposes?
[11:04am] welterde: sigblips: yeah.. that's the plan
[11:05am] sigblips: welterde: Thanks, I'll let you know later when I'm working with CUDA.
[11:05am] welterde: sigblips: storage-wise.. http://pics.welterde.de/hw/stuff/PICT6103.JPG <- IBM DSsomethingicantremember (2x17TB connected via 2x8GBit fibre channel to the box below the storage, which is then connected to the 40 gig infiniband fabric)
[11:06am] sigblips: jrseti_: We were talking about CAVE and CUDA earlier. You can read about it in the logs if you like. http://irc.sigblips.com/setiQuest/2012/
[11:06am] gliese581c: Title: Index of /setiQuest/2012 (at irc.sigblips.com)
[11:06am] sigblips: welterde: That RAID array must be really fast.
[11:07am] welterde: sigblips: and that's the FRAVE btw.. CAVE is something else.. that's located below the LRZ building if I recall building(LRZ is the local university DC.. top4 supercomputer and all ;)
[11:08am] lnigra joined the chat room.
[11:08am] sigblips: jrseti_: Do you know how fast the RAID array in the PowerEdge C2100 at Hat Creek is? http://setiquest.org/wiki/index.php/Hat_Creek_server_room#Hardware
[11:08am] gliese581c: Title: Hat Creek server room - setiquest wiki (at setiquest.org)
[11:08am] welterde: sigblips: indeed :)
[11:08am] jrseti_: No
[11:09am] sigblips: Welcome Lou. You can see what you've missed so far here http://irc.sigblips.com/setiQuest/2012/setiquest.09-25-2012.log We probably should start the meeting now.
[11:09am] lnigra: thanks
[11:10am] jrseti_: http://setiquest.org/forum/topic/community-meeting-2012-09-25
[11:10am] gliese581c: Title: Community Meeting 2012-09-25 | setiQuest (at setiquest.org)
[11:10am] jrseti_: 1) New AC is installed and working!
[11:10am] jrseti_: But...
[11:10am] lnigra: oh...
[11:10am] eb3frn joined the chat room.
[11:10am] jrseti_: Last night we could not get SonATA to run correctly.
[11:11am] jrseti_: After all went to bed we figured out the problem was one of the computers thought the year was 2008
[11:11am] jrseti_: The threw off all of our antenna positioning
[11:11am] jrseti_: We have fixed that, and I think it will all run tonight
[11:12am] sigblips: Bad battery?
[11:12am] jrseti_: I was at the ATA last week. The AC keeps th room really cold.
[11:12am] sigblips: What do you have the thermostat of the air conditioner set to?
[11:12am] jrseti_: We should be able to soon begin test using more of the Dell computers to get a wider bandwidth
[11:12am] welterde: . o O ( what? not running ntpd? ;)
[11:12am] sigblips: Welcome eb3frn. You can see what you've missed so far here http://irc.sigblips.com/setiQuest/2012/setiquest.09-25-2012.log
[11:12am] jrseti_: The AC is blowing out 66 degrees.
[11:13am] sigblips: Brrrrrrrrr.
[11:13am] jrseti_: This computer was not running ntp. It was newly set up and the person who did so neglected that
[11:13am] jrseti_: Must have been a battery problem. the computer was turned off for 3 weeks
[11:14am] lnigra: Will charlie be up? like to test fake setilive followups (hopefully) one last time
[11:14am] jrseti_: They are addressing the ntp issue - "should" not happen again!
[11:14am] jrseti_: Anyway, I have high hopes for tonight
[11:15am] jrseti_: "Charlie" should be running.
[11:15am] lnigra: excellent
[11:15am] jrseti_: ALL: FYI: Charlie is the server process that gets the signal reports from SETILive
[11:16am] lnigra: another question? any reason to expect all three beams up?
[11:17am] jrseti_: I am hoping 3 beams for tonight. But there is a problem - I think unrelated to the problem we had before the shutdown. So, tonight "may" be 3 beams.
[11:17am] sigblips: Was one of the beamformers acting up a couple weeks ago?
[11:17am] jrseti_: Yes, the last week we only used 2 beams.
[11:18am] jrseti_: That issue has been addressed.
[11:18am] sigblips: Just curious. What was the problem?
[11:19am] jrseti_: The network port stopped functioning on one of the BEE2s.
[11:19am] sigblips: OK, thanks.
[11:20am] lnigra: Yes, thanks for the update
[11:20am] jrseti_: The observing should start at around 7pm tonight.
[11:20am] jrseti_: fingers crossed!
[11:20am] jrseti_: Shall we move on to 2)?
[11:21am] jrseti_: I am building another SonATA distro. Or at least trying
[11:21am] jrseti_: SonATA was working well on 11.3
[11:21am] jrseti_: But when i tried to build a new distro, SonATA will not run
[11:22am] jrseti_: Severalof the links to packages that need to be installed have changed
[11:22am] jrseti_: I found the packages at other sites. All builds. But the test with the Voyager data does not run
[11:22am] jrseti_: So I am trying to figure out what to do
[11:23am] jrseti_: not fun!
[11:23am] jrseti_: I want to build a virtual machine with EVERYTHING installed.
[11:23am] sigblips: What is the Linux distro and version you were trying?
[11:23am] jrseti_: 11.3 OpenSUSE
[11:24am] jrseti_: That is what we are running at the ATA
[11:24am] jrseti_: But on real computers - not virtual ones
[11:24am] jrseti_: Still - having trouble with the DX nor connecting
[11:25am] sigblips: I thought that you had it working on your Mac with VMware Fusion last year?
[11:26am] jrseti_: I did! But I am trying to create a fresh version, a clean OS install.
[11:26am] jrseti_: I may have to take that distro that was working last year - copy it - then clean it up.
[11:27am] jrseti_: That was with the latest source code from last year. Hopefully if i do a code update from GitHub it will still work on that distro
[11:28am] jrseti_: Now I've talked my self into doing that.
[11:28am] sigblips: Have you checked that the Linux kernel versions are the same? Just to be able to rule that out.
[11:28am] jrseti_: good idea. Do they change kernel version with an openSUSE release number?
[11:29am] sigblips: Updates do and sometimes things break in the Linux kernel, subtle things.
[11:30am] jrseti_: That is what it seems like.
[11:30am] jrseti_: So I'll try that and report back next week
[11:30am] jrseti_: 3) Discuss the future of this IRC chat. See http://setiquest.org/forum/topic/future-irc-meetings
[11:30am] gliese581c: Title: future of IRC meetings | setiQuest (at setiquest.org)
[11:31am] jrseti_: If you guys feel we should continue, I am all for it.
[11:31am] sigblips: So far, 0 votes for Yes and 8 votes for No. Dave is the only person who commented in the thread.
[11:32am] sigblips: What are the SETI Institute's thoughts?
[11:32am] jrseti_: I was asked by the "boss' why we are still having the IRC chats.
[11:32am] welterde: sigblips: "yes" as in continue or cancel?
[11:33am] lnigra: "yes" was cancel, right?
[11:33am] sigblips: Yes = cancel. http://setiquest.org/content/should-weekly-community-irc-meetings-be-can...
[11:33am] gliese581c: Title: Should the weekly community IRC meetings be canceled? | setiQuest (at setiquest.org)
[11:33am] welterde: ah.. good :)
[11:34am] jrseti_: It is a good venue to help update about what is going on with SonATA
[11:34am] sigblips: Yes it is.
[11:36am] lnigra: Speaking for SETILive, I think there are still possibilities for collaboration. We hope to re-energize the site in the coming weeks.
[11:36am] sigblips: What are the SETI Institute's concerns and expectations about the future of the weekly community IRC meetings?
[11:36am] jrseti_: the concern was that it is a time sink.
[11:36am] sigblips: Inigra: This IRC channel could be used for that too.
[11:37am] jrseti_: Yes, a SETILive presence would be good on this IRC chat
[11:38am] jrseti_: lnigra: I'll try to email you to remind you each Tuesday
[11:39am] sigblips: I have it set as an automatic reminder on my phone.
[11:39am] jrseti_: me too
[11:39am] jrseti_: lnigra: can you give us a SETILive update?
[11:39am] lnigra: thanks, jrseti. My smartphone calendar should also help wean me of the need for reminders.
[11:40am] welterde: sigblips: and if you(or someone else) could highlight me before the meeting starts I would be most grateful(as I always forget ;)
[11:40am] lnigra: sure
[11:40am] sigblips: OK
[11:40am] lnigra: Ok. SETILIve: followups...
[11:41am] lnigra: fully expect fake triggered followups to generate a full sequence of 10 followups tonight. worked on live emulator.
[11:43am] lnigra: still have bug fixes that haven't been pushed live for rapid-fire followups triggers and confusion with the target-on/off status being common with the dx's
[11:44am] lnigra: the live detection algorithm basically works, but I think there's a bug that can cause a trigger in inappropriate conditions
[11:44am] lnigra: once that's fixed. live followups will be there.
[11:44am] lnigra: expected later this week.
[11:45am] jrseti_: great!
[11:45am] lnigra: We're also building a new version of the site which should be more engaging and easier to deal with those annoying RFI signals :-)
[11:46am] sigblips: What do you mean by live detection algorithm? The code that determines if enough people have classified something for a followup to happen? Or is there some new signal detection code in SETI Live?
[11:47am] lnigra: The former, sigblips: It's the same detection code - a nearest-neighbor grouping thing that determines if enough people agree well enough on a position and slope in a single beam.
[11:48am] lnigra: Stuart Lynn developed it. I think there's a simple bug in how the confidence in the agreement is quantified.
[11:48am] lnigra: Stuart has demonstrated that it basically works.
[11:49am] sigblips: OK, thanks.
[11:49am] lnigra: One technical feature in the new site will be polarization combining in the waterfalls.
[11:50am] lnigra: We've been serving the L and R polarizatiions separately and unless you have a gazillion people classifying, it doesn't make as much sense.
[11:50am] lnigra: It should reduce the noise background and reduce fading
[11:50am] sigblips: That's good. The two polarizations are often very similar.
[11:52am] lnigra: Another focus on the new site is to speed up the delivery of live data. You could wait up to 20 seconds sometimes.
[11:52am] lnigra: That's about it.
[11:52am] jrseti_: Thanks!
[11:53am] lnigra: Oh yes...
[11:53am] jrseti_: It will be great when the wait problem is solved.
[11:54am] lnigra: As soon as the new version is up. I'll immediately start looking into open API for our waterfall data at least and more if possible.
[11:56am] sigblips: Some people have expressed an interest in that. They'll be happy. It will be interesting to see what they create.
[11:57am] jrseti_: I'm hoping!
[11:58am] lnigra: Yes. I have hopes of making some tools available for regular users to investigate the RFI they find in their favorite signals, for instance.
[11:58am] jrseti_: Time is almost up. Anything else?
[11:59am] lnigra: nope
[11:59am] jrseti_: We now have a Donations$ link on the setiquest front page that should work
[11:59am] lnigra: You need a super pac sponsor
[11:59am] sigblips: It was setistars.org, the Donate buttons still are. Did setistars.org stop working?
[12:00pm] jrseti_: I really don't know what happed to SETIStars! I think the developer did not pull through
[12:00pm] jrseti_: what donation button?
[12:01pm] sigblips: The yellow ones.
[12:01pm] jrseti_: Oh! I'll change that! Thanks!
[12:02pm] jrseti_: I'll have to figure out how to change that
[12:02pm] sigblips: I regularly check setistars.org, they've been averaging about a donation a week. So it still is making some money for the SETI Institute.
[12:03pm] jrseti_: Sigblips: do you know how to edit that button?
[12:03pm] jrseti_: Can it be done through the admin part of the site?
[12:03pm] sigblips: No I don't. I think you need to dive deeper into Drupal than you can with the Administration screen.
[12:04pm] sigblips: If I see it I'll let you know. You'll probably have to ssh in and dig around.
[12:04pm] jrseti_: I'll take a look.
[12:06pm] jrseti_: I did, and if I grep ALL the files, "Donate" does not appear in any
[12:06pm] sigblips: It has to be somewhere.
[12:07pm] jrseti_: So it must have to be done in the admin interface. I think Drupal uses a database where things hide
[12:08pm] lnigra left the chat room.
[12:08pm] sigblips: The setiquest.org menus are different than what is in the Admin interface. I think the Drupal developer hard-wired a bunch of things during the last re-design with Avinash.
[12:08pm] jrseti_: Looks like it!
[12:08pm] jrseti_: I'll have to rummage around in the database
[12:09pm] sigblips: I think Michael fixed the Donation button links many months ago.
[12:10pm] jrseti_: I'll email hime
[12:10pm] jrseti_: Shall we end the meeting?
[12:10pm] sigblips: Same time, same channel next week?
[12:11pm] jrseti_: yes. Thank you!
[12:11pm] sigblips: Bye all. I'll post the transcript in the setiQuest forum.