- codes-ross-users - mailman3.cels.anl.gov

updates on workload support
by Jenkins, Jonathan P. 10 Dec '14

10 Dec '14

Hi all, I’ve had a bit of time over the past few days to make some progress on the IO/network workloads APIs of CODES. In particular, I have experimental branches on both codes-base and codes-net repositories (both named workload-combine) that integrate the network workload API and parsers into the IO workload API. In other words, the directory “codes-base/src/network-workload” is now a dead directory, with relevant content having been moved to “codes-base/src/workload”. At some point this will be pushed to master, but there are some folks who are using the current setup on master that may need to put in fixes or other changes, so I’m holding off on merging for the time being. Some notes: - the scalatrace workload generator is only a partial implementation, and won’t work for simulations (it assumes one reader per MPI rank, whereas multiple LPs can be calling into workload functions on the same MPI rank). The dumpi generator, another MPI tracer (http://sst.sandia.gov/about_dumpi.html) is more complete for this purpose. - the existing workload generators need some cleaning/refactoring. They currently are not too pretty. - I also need to update the documentation, but the usage is near identical to the original IO workload API. - there aren’t any generators yet that generate both IO and network events (contributions here would of course be welcome :) ). Thanks, John

1 0

RNG behavior change in codes-base master
by Jenkins, Jonathan P. 08 Dec '14

08 Dec '14

Hi all, FYI, pretty soon I’ll be pushing a change to the codes-base repository that changes the random number generator (RNG) behavior for models using CODES. In CODES, we have a heavily-used function called codes_local_latency that adds a tiny amount of random noise to event timestamps to help prevent event ties and keep parallel simulations deterministic. Previously, we were using it with the default RNG (i.e. lp->rng[0]. BTW, ROSS allows multiple random number generators for each LP. Cool!!!). However, we'd like to insulate the RNG sequences that “matter” (those called by the user) from those that don't (our noise function), to improve model determinism in the face of minor changes that affect the number of RNG functions called. No end-user code needs to be updated, but model behavior that depends on RNG will change. Thanks, John

1 0

CODES 0.3.0 released!
by Jenkins, Jonathan P. 07 Nov '14

07 Nov '14

Hi all, CODES 0.3.0 is released! This marks the first formal release of the codes-base and codes-net projects. You can find the release at http://mcs.anl.gov/projects/codes/downloads. Against recent repository revisions, this release includes more complete and cleaned up documentation. Thanks, John

1 0

Getting Started Guide
by Joe Scott 29 Oct '14

29 Oct '14

Hello, I am evaluating CODES for building a high level simulator of some of the architecture changes envisioned for PVFS3. I found the latex docs in the source tree, and I think I have somewhat of an understanding of how CODES is built. A couple questions: Are there plans to update/complete the documentation in the near future? Is CODES still "alive" enough to start building on? Thanks, Joe Scott Clemson University

2 1