CAWCR-BoM ACCESS NWP Ngamai Migration Working Group

Issues and Task List table

  • Updated after Meeting 5, Wed 11am, 7th August, 9E Meeting room.
  • Changes/updates to previous notes are in bold.
  • This table contains the main current items relevant to the working group.
  • Previous versions of this table are also available in notes for Meetings 1-4; these can be accessed from links on the main page.
  • Issues no longer active can be seen in earlier table versions.

3 Setup and rebuild of /apps Ongoing work. sms and python almost done. perl which is required for fcm2.3 to be done. fcm1.5 works with std perl.
ACTION: Monitor Status.
/apps, rab
6 Intel compiler 11.0.83 * Details moved to NewSun_IntelCompiler_11.0.83
* Re-download & rebuild of v11 will be attempted by Tim Pugh and Justin Freeman. If successful, will be compatible with openMPI, will not need MPI/Sun.
* Questions: Will single executable built with identical compiler options, using non-machine specific setting, identical compiler version, openMPI library instead of MPI/Sun, give bit identical result on raijin-ngamai-solar.
* OpenMPI1.6.3/5 at NCI is built using GNU compiler. At BoM, Intel Compiler was used - what is the impact?

* Robin update: info from Tim Pugh; Justin to try doing this, but he's currently on leave for 2 weeks.
* Importance of this as a fall-back if there are problems with v12 execs was again emphasised.

ACTION: Continue investigations.
/apps, rab
7 Source Code for UM No action required. Original repository already in access-svn. azs
8/9/10 Source Code for VAR,OPS,SURF In general, migrate all SVN repositories from solar early October 2013. Use accessdev if it is ready. Otherwise migrate to ngamai first (without access from nci). Ask for increase in diskspace for access-svn.
ACTION: Prepare for migration. Monitor reliability of access-svn server.
11 ~access directory Goal of having interoperability with raijin is not straightforward, particularly with executable and libraries. Need to handle case by case. $HOST or $MACH subdirectory required in some instances. Also see Scott's tildeAccess notes.
ACTION: Address case by case as ~access is being setup. Investigate transfer speed solar-ngamai.
12 GCOM Libraries Updated path/name: example: ~access/apps/gcom/GCOM3.5/ngamai/bld_12.1.8.273_1.6.5_new_optns-03.
ACTION: Information on building gcom to be documented on the wiki.
13 Migrate Trac databases Cutover from solar to ngamai 1st weekend of Oct.
ACTION: Prepare migration plan.
14 UM Small execs Do for vn7.3, 7.5, 7.6, 8.2 and 8.4. May be able to use copy from raijin.
ACTION: No report.
15 Migrate UMUI, VARUI, OPSUI and SCSUI Initially to operate UMUI from solar. Big Font issue encountered on ngamai. umuix which uses ghui3.0 does not have the problem, but it is desired that both ghui2.0 and ghui3.0 to work before migrating.
ACTION: Workaround exists, but the earlier the migration the better.
azs, zhihong, ilia, xiao, say
16 CAP program on ngamai vn8.1 now available and set up on Raijin. Sufficient for time being; not urgent to port to ngamai, as we can create ancils on raijin and copy to ngamai.
ACTION: No action required.
17 Re-compile/build UM 7.5/7.6 Executables for APS1 - Global, Regional, Access-C ... * Documentation of build procedure, UI job id, wiki page of build or similar to UKMO Parallel suite page.
* Sample of build documention to be created for discussion.
* EXAMPLE FOR REF, see UKMO Parallel Suites documentation on collab wiki.
ACTION: Detailed documentation for each build should be available on wiki.
--17a ACCESS-G * VN7.6 (PS25)
* Custom-built v11 execs reproduce AG1 on solar.
* v12 execs built and tested to give similar results on solar & ngamai, in use in Xiao's ngamai AG1 test NWP suite.
* Zhian's mods also included.
* Working on producing documented build procedure.

* See AccessGBuild for more details.
--17b ACCESS-R/R12 * VN7.6 (PS25)
* Custom-built v11 execs reproduce AR1 on solar.
* v12 testing on ngamai will follow AG1.
--17c ACCESS-C * VN7.6 (PS25)
* Custom-built v11 execs reproduce AC1 on solar.
* v12 execs built and tested to give similar results.
* Working on producing documented build procedure.

* See AccessCBuild for more details.
--17d ACCESS-TC * VN7.6 (PS25) ATC1 will use same execs as AR1. ilia,xiao,azs
--17e UKV VN7.6 (PS25) ilia,xiao,azs
--17f VAR * Compile/build of VAR not done through UI.
* Recompiled VAR gives different result.
* -model_precise compile option is important
--17g OPS Compile/build of OPS not done through UI ilia,xiao,joan
--17h SURF Compile/build of SURF not done through UI. ilia,xiao
--17i Martin's Worms Diff plots, ensemble plots, etc. will be useful for verification.
ACTION: Get update from Martin at next meeting. See APS1G_rmsdifferences for an example
18 MARS scripts and exes Archiving scripts, converters etc.
ACTION: Get update from Tan Le at next meeting.
19 Porting/Migration website * Use NCI access trac.
* Need NCI userid for everyone.
* Need more disk space & proper backup regime for nci.trac server.
ACTION: Monitor status.
20 Web serving and documentation from ngamai httpd from ngamai is active. nothing setup yet.
ACTION: Monitor status.
rab, Porting Group
21 fcm utility fcm 1.5 used succssfully, but not fcm 2.3. fcm2.3 will need perl with appropriate libraries from /apps. See 3.
ACTION: Monitor status.
martin, /apps
24 UM Ancillaries input data. Since UM ancillaries, in particular for N512 can be quite large, propose that it go into /access/ancils instead of into UMDIR (or into /access/data/ancils).
ACTION: No update.
26 APS1 suites * Xiao has working ngamai AG1 NWP suite based on NMOC SCS components; using v12 UM/VAR/OPS executables from 17. Has successfully run 2 cycles from 21 July. Results are different from solar, based on residuals in first VAR step; don't have convenient diagnostic for OPS step. Concerned that VAR residual differences are bit larger than expected.
* Planning to start one-month test .

ACTION: Chris/Xiao/Zhihong to look at OPS/VAR component assessments.
27 APS2 suite APS2 porting is of secondary importance since it is not operational yet.
ACTION: None for now.
34 Higher management's Porting plan. * Robert Jukic appointed as Project Manager by Steering Commitee.
* Single-page timeline has been produced identifying major steps for CAWCR and NMOC.
* More detail on these steps and resource reporting will be developed further.
ACTION: Robin will draw up draft, discuss with individuals responsible for each of the work items.
35 SMS * To be installed as a "module", as part of /apps.
* New NMOC included files from Milton Woods.
* CAWCR research may stay with their own set of include files, e.g. to put "SMSOUT" into different directory.
ACTION: Update documentation on wiki page
36 Verify * NWP verification software to be ported to ngamai and raijin.
* Chris Bridge to handle this, NMOC to manage.
* ngamai version will be included in /apps; raijin version to follow.
ACTION: NMOC to report progress on this item.
37 Python-matplotlib plotting Systems Wenming to port to ngamai, raijin. Requires python-matplotlib. wenming, /apps
38 Configuration Management of systems/suites that go into operation Discussion on this subject moved to APS_ConfigurationManagement. Porting Group
39 Turboboost * Turboboost is set on raijin.
* Not critical to operational porting activity.
* Will be investigated on ngamai after porting has been completed.
40 Hyperthreading * Not critical to operational porting activity.
* Will be investigated on ngamai after porting has been completed.
41 Level of thread support for OpenMPI library * Not critical to operational porting activity.
* Will be investigated on ngamai after porting has been completed.
Last modified 4 years ago Last modified on Oct 20, 2016 2:35:47 PM