[Shake-dev] ShakeMap and Earthworm

Thu Jul 19 15:45:08 UTC 2012

Bruce:

Ah ok. AQMS is a DB driven version of EW. I do not use it at the moment.

I know how to proceed then. I will write a python script to check for new
events in the /data folder and then call ./shake <event_id> and put the
script in a crontab.

Let me say again that you have been extremely helpful. I will keep on eye
on the list and contribute when and where I can.

Kind Regards,

*Branden C. Christensen*
Director, OSOP
*branden.christensen at osop.com.pa*
www.osop.com.pa
Volcan, Panama

On Thu, Jul 19, 2012 at 3:32 PM, Bruce Worden <cbworden at gmail.com> wrote:

> Hi Branden,
>
> queue was designed to work with systems running the AQMS software. There
> are two programs (shake_alarm and shake_cancel) that are run (in an AQMS
> system they are started by database triggers, but you can also run them
> manually) to tell queue about a new, updated, or cancelled event. queue
> then accesses the AQMS database to get event info and decide whether to run
> the event and what priority it should be given. queue also allows you to
> configure automatic repeat runs, different magnitude thresholds for
> different geographic regions, etc.
>
> But since you're not running AQMS, you'd have to modify it. There is a
> program called queue_qdds that was triggered by events coming into the old
> QDDS system. That might be an easier target for modification. It's not my
> program, but I'm guessing you could get it to work.
>
> Either way, though, you will still need some kind of system to recognize
> new events and either kick off shake, or poke queue.
>
> I don't know what other earthworm users have done in the past, but the one
> user that I think might be able to help you is out of the office until the
> 30th.
>
> Bruce
>
>
> On Jul 18, 2012, at 6:45 PM, Branden Christensen wrote:
>
> Bruce:
>
>
> Last step here.
>
> I am looking to automate the production of shakemaps. It looks like I
> might have 2 paths:
>
> 1. Write a script that looks for new event folders and executes the shake
> -event # command and then put the script in the crontab
>
> OR
>
> 2. use queue? e.g., ./queue&
> Will queue really work? What is the story with queue? Is it meant to run
> in the background like this? The manual suggests as much.
>
>
> Best,
>
>
> *Branden C. Christensen*
> Director, OSOP
> *branden.christensen at osop.com.pa*
> www.osop.com.pa
> Volcan, Panama
>
>
>
>
>
> On Tue, Jul 17, 2012 at 11:05 PM, Bruce Worden <cbworden at gmail.com> wrote:
>
>> You shouldn't serve the web pages directly from the events' data
>> directories -- actually that won't work because a lot of stuff will be
>> missing. That's where 'transfer' (or your version of it) comes in. You set
>> it up so that you serve from "<some_directory>/shake". Then 'transfer'
>> sticks the new index page there along with stuff in "archive" (overwriting
>> anything that was there before) and the new event's directory. Each new
>> event needs to update the index page and the archive, but I agree that the
>> way it is handled leaves a lot of junk lying around in the event's working
>> directory.
>>
>> Before the site will work, you need to configure web.conf do 'make web,'
>> then go to bin and run ./transfer -permweb (see p. ~12 of the software
>> guide for this step). transfer -permweb currently doesn't work, but there
>> will be a fixed version in the repository by the end of today. You'll do
>> "svn update" and "make" to get the new code. There will be a number of
>> other fixes, too. Remember to protect your custom version of topo2grd
>> before updating.
>>
>> You may also want to go to <shake>/lib/genex/web/config and edit
>> archivepages.xml to remove all of the years for which you don't have
>> earthquakes.
>>
>> These are old-style web pages that use almost no modern web technology.
>> They aren't particularly pretty, and we don't really do much to update
>> them. The USGS pages are slick and modern, but nobody has template-ized
>> them yet, so users are either stuck with what we distribute, or they can
>> put together their own.
>>
>> Bruce
>>
>>
>> On Jul 17, 2012, at 2:11 PM, Branden Christensen wrote:
>>
>> I have noticed something off though. If you create a directory for EACH
>> event then there will be gazillions of websites:
>>
>> e.g.,
>>
>> http://sandbox1.osop.com.pa/shakemap/9583161/genex/web/shake/
>> http://sandbox1.osop.com.pa/shakemap/4035/genex/web/shake/
>> ...
>> ...
>> ..
>> where each event has the directory structure of the entire website:
>>
>> e.g.,
>> osop at sandbox1:/opt/shakemap/data$ ls 9583161/genex/web/shake/archive
>> 1999.html  2002.html  2005.html  2008.html  2011.html   major.html
>> 2000.html  2003.html  2006.html  2009.html  2012.html   pre1999.html
>> 2001.html  2004.html  2007.html  2010.html  index.html  scenario.html
>> osop at sandbox1:/opt/shakemap/data$ ls 4035/genex/web/shake/archive
>> 1999.html  2002.html  2005.html  2008.html  2011.html   major.html
>> 2000.html  2003.html  2006.html  2009.html  2012.html   pre1999.html
>> 2001.html  2004.html  2007.html  2010.html  index.html  scenario.html
>>
>>
>> *Branden C. Christensen*
>> Director, OSOP
>> *branden.christensen at osop.com.pa*
>> www.osop.com.pa
>> Volcan, Panama
>>
>>
>>
>>
>>
>> On Tue, Jul 17, 2012 at 9:04 PM, Bruce Worden <cbworden at gmail.com> wrote:
>>
>>> Nice. Thanks for letting us know. Now you just need to get the files
>>> dumped into directories "<evid>/input", kick off shake, and you're good to
>>> go.
>>>
>>> Bruce
>>>
>>> On Jul 17, 2012, at 1:20 PM, Branden Christensen wrote:
>>>
>>> Hi all:
>>>
>>>
>>> to generate  _dat.xml files: gmew
>>> to generate event.xml files: localmag
>>>
>>> :)
>>>
>>> e.g.,
>>>
>>> ls /opt/shakemap/data:
>>> ...
>>> 4034_0_event.xml
>>> 4034_1_event.xml
>>> 4034_2_event.xml
>>> 4034_dig_dat.xml
>>> 9583161
>>> tmp
>>>
>>>
>>> *Branden C. Christensen*
>>> Director, OSOP
>>> *branden.christensen at osop.com.pa*
>>> www.osop.com.pa
>>> Volcan, Panama
>>>
>>>
>>>
>>>
>>>
>>> On Fri, Jul 13, 2012 at 6:48 PM, Branden Christensen <
>>> branden.christensen at osop.com.pa> wrote:
>>>
>>>> Hi all:
>>>>
>>>>
>>>> I am looking to get my ShakeMap installation (thanks Bruce) to work
>>>> with the output from Earthworm's gmew.
>>>>
>>>> At the moment I have gmew configured to write to <shakemap_home>/data:
>>>>
>>>> osop at sandbox1:/opt/shakemap/data$ ls
>>>> 3739_dig_dat.xml  3743_dig_dat.xml  3744_dig_dat.xml  9583161
>>>>
>>>> From the manual and the example "9583161" I see that gmew is not doing
>>>> everything I need. I appear to be lacking the formation of a directory
>>>> specific to EACH event and an event.xml file.
>>>>
>>>> Your insights are very much appreciated.
>>>>
>>>>
>>>> Kind Regards,
>>>>
>>>> *Branden C. Christensen*
>>>> Director, OSOP
>>>> *branden.christensen at osop.com.pa*
>>>> www.osop.com.pa
>>>> Volcan, Panama
>>>>
>>>>
>>>>
>>>>
>>> _______________________________________________
>>> Shake-dev mailing list
>>> Shake-dev at geohazards.usgs.gov
>>> https://geohazards.usgs.gov/mailman/listinfo/shake-dev
>>>
>>>
>>>
>>
>>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://geohazards.usgs.gov/pipermail/shake-dev/attachments/20120719/8924d7aa/attachment-0001.html>