McMurdo Dry Valleys SiteBytes 2012

What has happened during the last year at McMurdo IM?

MCM entered a new funding cycle last year (MCM-IV). The MCM PIs decided before hand that the MCM-IM will be co-located with the SEV and the LNO, with the forward looking idea of leveraging resources and expertise. That served well! Just last week I found that I needed to install a "Windows Server" yesterday. Needless to say, an unused server appeared out of nowhere, and even a license for Windows Server. I never done that before, but I had helping hands (thanks Dez) to guide me through the ordeal.

New funding cycle, new location, new IM. Those three "news" summarize last year.

The big help came from Chris Gardner (big thumb up), Sue Welch and Tina Takacs-Vesbach

1) Get familiar with the IM-related protocols at MCM (read: get ready to change the tires while the car is driving at 60mph.) MCM IM speak:

MCM <-> IT World
----------------------
Database <-> Information management system
Database person <-> Sysadmin
Database person <-> Web mistress
Database person <-> Information manager
Database person <-> Database administrator
Database person <-> Data entry person
Database person <-> Houdini
You get the picture

2) Set up a Sun T3-1 server running Solaris 5.10 ( I have a blog with threaded posts under the header "Surviving Solaris - thanks Dez, Renee)
The lifesaver here is opencsw.com, the closest thing to Ubuntu's repo (apt-get install blahblah -- here is pkgutil -i blablah). T3-1 motherboard blew out on a friday :)

3) Migrate Java Server Pages from a 5 to 7 year old deploy ( old version of Java, old Tomcat - thanks Mark ) to the new server. Make Tomcat dance
with Apache (a connector). Chris helped a lot. How the hell does he remember details from 7 years ago?

4) Catch up with season data. Find out what is that the teams collect, who owes the IM team what, and ask (awesome MCM team, with established chain of command and high expectations on data sharing -- in general -)

5) Backups. Backups of Backups. And more backups. Tera-drives, LNO new server, other Ubuntu boxes. Mainly, I backup the database, and a hard code of parts of the code, and the whole site.

6) Update the "MCM IM book" with the new changes that pertain to the UNM.

7) Break Drupal to enhance the already rockin' functionality (the site has a data query system that allows you to specify a date range -- top that!)
Introduced Drupal Bibliography module (an estelar app!) and personnel management. Teased with maps. But before revolutionizing it all, better to gain trust -- deliver on the plans that set out to do. Too much to disclose here!

8) The EML. oh, the jolly EML. Turn 189 instances of EML2.0.1 into EML2.1.x w/ further checks (it is a new era indeed!). I used Perl to do the following changes:
8.a) 2.1.0 compliant -- run stylesheet
8.b) metadataProvider, contact, publisher - from "Byrd Polar ctr" to UNM.
8.c) packageID Identifier from 000XX to XX (per Duane's guidance)
8.d) package revision from Y to Y+1
8.e) copy distribution URL at "resource" level into dataTable/physical/distro need be - watch with existing ones
8.f) add broad keywords (McMurdo, ..) to all EMLs
8.g) remove the tags surrounding the body. A workaround to a serious bug on metacat.
8.h) make sure specified the exact header of the columns of the data described.

If you can use some of that, get in touch with me.

9) Amazon Web Services. Amazing stuff.
9.a) Use the bucket (S3) to store some backups and also store static web content. Us
9.b) Use the Elastic Instances (EC2) to deploy a large Ubuntu instance with one click.
9.c) Use the Relationa Databases (RDS) to deploy an Oracle enterprise BYO to host an MCM replica.

The idea is to have an off-location instance that can pick up the traffic for MCM when the UNM's internet access is cutoff (too windy, a small flood, a electrical storm, maintenance, a human screw-up or combos of those factors). I still need to connect Apache with tomcat and the DB, move the direct connection in EC2 to the bucket (saves $$) and negotiate agile changes of DNS entries to redirect traffic as needed.

10) A new section on the History of MCM. Our sociologist has stories to tell that make a fun read. See for example his latest piece on the LTER Newsletter on "Taylor Valley's missing lake".

11) Plans for a new section on microbial diversity. It is the molecular level. That is where we see biodiversity at a scale that dwarfs our knowledge. yet, genomics is not in Kansas anymore for the casual ecologists. It presents new challenges for the information manager. If we only have a process where EML or us were more dynamic, may be LTER can actually lead a bit in this exploding field. Oh, I was saying, Plans for a new section on microbial diversity. We looked at Wade's Sapelo. NTL's clone. Now we look at the new (old) boys on the block (MG-RAST). We ask what we want to see. We plan to deliver. Many fronts: The genomics standard consortium, the Biodiversity working group, the microbial observatories, the AntaBif (an IPT incarnation of GBIF for antartica, through a Terrestrial Observatory Network championed by B. Adams

12) Sanity/Project management made simple: an IM "to do" list ( at this address - deims.lternet.edu/mcmdemo )

13) Review the IM-related comments that steamed from the last site review -- address them (how fun)

14) Place some datasets on PASTA. How about 60?

14) Put out fires.
14.a) I cannot find XYZ data. Help.
14.b) Install this software. Help.
14.c) The data ABC is wrong, take it down. Help.
14.d) At Sci Council is reported that MCM EML needs help (is crap). OMG.
14.e,f) ......

...To be continued....