search

codes-org / codes

The Co-Design of Exascale Storage Architectures (CODES) simulation framework builds upon the ROSS parallel discrete event simulation engine to provide high-performance simulation utilities and models for building scalable distributed systems simulations
Other
40 stars 16 forks source link

"Error: failed to find workload generator dumpi-trace-workload" #207

Closed lwh1990 closed 3 years ago

lwh1990 commented 3 years ago

Hi, I am trying to run the following example shown below and I am using the master branch for codes and ROSS.

./model-net-mpi-replay --sync=1 --disable_compute=1 --workload_type="dumpi" --num_net_traces=27 --workload_file=df_AMG_n27_dumpi/dumpi-2014.03.03.14.55.00- -- ./src/network-workloads/conf/modelnet-mpi-test-torus.conf

But it show some errors:

Fri Aug 7 03:52:27 2020

ROSS Version: HEAD-HASH-NOTFOUND

tw_net_start: Found world size to be 2 NIC num injection port not specified, setting to 1 NIC seq delay not specified, setting to 10.000000 NIC num copy queues not specified, setting to 1 Bandwidth of compute node channels not specified, setting to 20.000000 within node transfer per byte delay is 0.050000 Within-node eager limit (node_eager_limit) not specified, setting to 16000

ROSS Core Configuration: Total PEs 2 Total KPs [Nodes (2) x KPs (16)] 32 Total LPs 64 Simulation End Time 3600000000000.00 LP-to-PE Mapping model defined

ROSS Event Memory Allocation: Model events 8193 Network events 16 Total events 8208

Error: failed to find workload generator dumpi-trace-workload Error: failed to find workload generator dumpi-trace-workload Error: failed to find workload generator dumpi-trace-workload Error: failed to find workload generator dumpi-trace-workload Error: failed to find workload generator dumpi-trace-workload Error: failed to find workload generator dumpi-trace-workload Error: failed to find workload generator dumpi-trace-workload Error: failed to find workload generator dumpi-trace-workload Error: failed to find workload generator dumpi-trace-workload Error: failed to find workload generator dumpi-trace-workload Error: failed to find workload generator dumpi-trace-workload Error: failed to find workload generator dumpi-trace-workload Error: failed to find workload generator dumpi-trace-workload Error: failed to find workload generator dumpi-trace-workload Error: failed to find workload generator dumpi-trace-workload Error: failed to find workload generator dumpi-trace-workload Error: failed to find workload generator dumpi-trace-workload Error: failed to find workload generator dumpi-trace-workload Error: failed to find workload generator dumpi-trace-workload Error: failed to find workload generator dumpi-trace-workload Error: failed to find workload generator dumpi-trace-workload Error: failed to find workload generator dumpi-trace-workload Error: failed to find workload generator dumpi-trace-workload Error: failed to find workload generator dumpi-trace-workload Error: failed to find workload generator dumpi-trace-workload Error: failed to find workload generator dumpi-trace-workload Error: failed to find workload generator dumpi-trace-workload START PARALLEL OPTIMISTIC SIMULATION WITH SUSPEND LP FEATURE

model-net-mpi-replay: src/workload/codes-workload.c:237: codes_workload_get_next: Assertion tmp' failed. model-net-mpi-replay: src/workload/codes-workload.c:237: codes_workload_get_next: Assertiontmp' failed. [localhost:29913] Process received signal [localhost:29912] Process received signal [localhost:29913] Signal: Aborted (6) [localhost:29913] Signal code: (-6) [localhost:29912] Signal: Aborted (6) [localhost:29912] Signal code: (-6) [localhost:29913] End of error message [localhost:29912] End of error message

Primary job terminated normally, but 1 process returned a non-zero exit code. Per user-direction, the job has been aborted.

Thanks for any help.

lwh1990 commented 3 years ago

The problem solved after I installed the dumpi library. This issue can be closed.

nmcglo commented 3 years ago

Glad you were able to figure it out!

lzk23 commented 1 year ago

Hi lwh1990， This problem also occurs in my test. I have installed the dumpi library as you suggested. However i don't know how to link CODES to DUMPI. Could you specify it. Thanks.

lzk23 commented 1 year ago

Hi nmcglohon, Could you specify how to link CODES to DUMPI library. Thanks

lzk23 commented 1 year ago

The problem occurs when i want to replay application traces on network models as the illustration at https://github.com/codes-org/codes/wiki/codes-dumpi-workload.

BingluCS commented 7 months ago

The problem occurs when i want to replay application traces on network models as the illustration at https://github.com/codes-org/codes/wiki/codes-dumpi-workload.

Hi lzk23 I also confront the same problem. Do you solve it?

nmcglo commented 7 months ago

There's some very old documentation for it somewhere in the wiki if you're using auto tools to build codes.

I converted codes to CMake a few years ago and there's a DUMPI_BUILD_PATH cmake config that you point to your dumpi installation directory.

It's been a while since I've worked with CODES though so I don't recall the details thoroughly, unfortunately.

BingluCS commented 7 months ago

Thanks, I have found the solution with your help. I used the option --with-dumpi=/path/to/dumpi/install with configure.