...

Summary information for Hoffmann

...

researchers

...

, as of

...

12/

...

12/13:

Helios and Sol will be unavailable to researchers between Monday, Jan. 6th at 9am and approximately Thursday, Jan 9th, at noon.

• Before the shutdown, all data of value must be copied off of Sol.
  • All user directory data will be removed.
• Before the shutdown, all jobs on both Sol and Helios must be stopped.
  • Any remaining jobs will be stopped by ChemIT staff
• Only Sol will be up after this work is done.(

...

• Helios will no longer be available after all this is done.)
  • Data will be moved from Helios to Sol.
  • And Helios's nodes will be added to Sol.

Detailed information for cluster migration:

1) Reduce the amount of researchers’ temporary, test user data on Sol to the bare minimum during the break.

• After the break, ChemIT will be deleting the content within everyone's test home directories on Sol. (Those directories We will be deleting all the home directories. (They were originally copied there as a testing convenience during our test phase.)
  • We do this deletion before the final copy to
  enable
  • ensure a clean import of
  researchers'
  • researchers’ current production home directories from Helios.
  • Lulu
  has
  • advised
  them
  • researchers of this
  from
  • step before the Thanksgiving break, so hopefully no surprises.
• During the winter break, Sol's data will be backed up.
  • As with Helios, Sol's data will be copied and versioned to a dedicated, internal hard drive.
  • NEW: EZ-Backup will be enabled on Sol, providing off-site backup protection.

Bottom line: If there are any files on Sol which researchers cannot afford to lose

...

,

...

researchers must move that data to Helios

...

(

...

or elsewhere) before Sol is turned off.

2)

...

Monday, Jan. 6th at 9am: Helios AND Sol being turned off to researchers

Researchers must know that:

• ALL jobs must cease by then. Any remaining jobs will be turned off by ChemIT.
• Also, there must be no remaining research data on Sol.

Tasks for ChemIT to do:

• Copy researchers' production home directory data from Helios to Sol.
• Move Helios's compute nodes into Sol's rack and reconnect all networking and power.
  • This allows us to add the older compute nodes from Helios, making Sol more powerful.
• Testing by ChemIT.
  • Coordinate with Huayun (Deng) for initial testing, perhaps sometime Wednesday.

3) Thursday, Jan 9th, at noon: Expect Sol available to researchers.

• At this point, Sol will be up in full production and open to all researchers, with some or all of Helios's compute nodes attached.

The Hoffmann group is encouraged to improve their software installations in order to create a more robust environment and to improve support outcomes.

Tasks for Hoffmann group:

• Confirm that Sol is "good". That is, there is no need to "roll back" to using Helios's head node and for us to start again.
• Hoffmann group members are encouraged to install their shared group software in the /home/hoffmann/bin/ directory. Including:
  • Materials Studio software. (Who?
• Sol will need to be turned off and worked on by us. (Approximate duration?) Therefore, it will be off-line to researcher.
• Ad hoc permission to select researches (Prasad?) will be granted so they may install and test their software.
• Tasks for us to do:
  • Get the server's versioning hard disk recognized within the new head node.
  • Lulu to install the Intel compiler in the /software/ directory, with the proper licensing.
  • Your group (Prasad?) to install the Materials Studio software in the /data/ Hoffmann/bin/ directory. If Prasad is not available to do this, let's let’s discuss our your options- thank you.
    

3) Researchers will be asked to test select software on Sol.

• Changes to where the Intel compiler and Materials Studio reside (to facilitated debugging and support) must be confirmed to work for researchers.
• Again we are testing on Sol. Thus, the only production work should still be on Helios, not Sol.

Oliver's quick notes:

11/15/13, from email sent to Roald, by Oliver:

• We are still working to learn how much of the Intel package components your group needs/ wants, and to what degree a non-current version might meet their needs in case we don't need to invest in the most recent version

The rest of the email:

I believe we have a short-term answer to address your group's need for the Intel compiler. This temporary solution will work as long as we are earnestly working on a long-term solution.

When you return, I would appreciate discussing options to meet your long-term needs cost-effectively. Michael Hint and I have some imaginative ideas, but some involve coordinating with other CCB researchers (Scheraga and Ananth).

Michael Hint learned the difference between a single user license and the 2-person concurrent licenses. The answer (below) informs one of our ideas, if you care to read it before we meet. But no problem if you don't get to it since I  can summarize when I discuss our ideas.

==================================

Sent: Friday, November 08, 2013 4:00 PM
To: Guoying Gao
Cc: Oliver B. Habicht
Subject: one computer matter which will come up

Hi, Guoying,

               In his conversation with me, Oliver said that he has investigated the error message referring to absence of a license for an Intel Compiler. It indeed appears that we had been using an older compiler, and that to use it in the future we will need to get a license. But that license is costly; Oliver is determining how much. What he will ask you, and you might inquire of the group in preparation is how often we use that compiler – daily, once a week, once a year? And how many people in the group are using it.

Thanks,

roald

• • )
  • vasp
  • gulp
  • etc.

Tasks for ChemIT:

• Once Hoffmann researchers confirm Sol is "good", Helios's head node can be added as a compute node to Sol.

NOTE: Helios will no longer be available after all this is done.

• Data will be moved from Helios to Sol.
• And Helios's nodes will be added to Sol.

Thank you -ChemIT

...

Older or other notes:

Data rates

12/11/13: Users data to transfer from Helios to Sol is about 80GB. The transfer time, using rsync, is expected to take about 24 hours.==================================

Next steps

• Meet to review all options and confirm desired direction and expected timing.
• Review resources. Huayun Gen has cluster management experience, including set-up.

...