Parallel Processing in EMAN2

EMAN2 uses a modular strategy for running commands in parallel. That is, you can choose different ways to run EMAN2 programs in parallel, depending on your environment. Unfortunately, as of April, 2010, there is still only one available parallelism strategy. This should be gradually fleshed out over 2010. Also unfortunately, it isn't trivial to use for simple multithreaded execution (but it does work). We hope to rectify this soon.

Programs with parallelism support will take the --parallel command line option as follows:

--parallel=<type>:<option>=<value>:<option>=<value>:...

for example, for the distributed parallelism model: --parallel=dc:localhost:9990

for the local multicore threaded model: --parallel=thread:4 (where 4 is the number of cores to use)

GPGPU Computing

While not precisely a parallelism methodology, this technique makes use of the GPU (graphics processing unit) common in most modern PC's, to dramatically accelerate many image processing algorithms. At present (summer 2009) we are at the initial stages of implementing GPGPU support using Nvidia's CUDA infrastructure. We will likely move to OpenCL in future as it becomes a stable platform. We have only implemented a few algorithms using this methodology to date, and we will need to implement and optimize virtually all of them before this becomes a viable platform for day-to-day use. However, we have demonstrated speedups of as much as 100x in select algorithms, meaning a desktop PC with a GPU could easily become the equivalent of a small Linux cluster. While all of the GPGPU code is available in the nightly source snapshots, you are encouraged to contact sludtke@bcm.edu if you are interested in experimenting with this technology.

Local Machine (multiple cores)

As of 7/15/2010 this is now supported ! If you only want to use multiple cores on your local machine, just put 'thread:<ncpu>' in the 'Parallel' box, or specify the '--parallel=thread:<ncpu> option on the command line. <ncpu> should, of course, be replaced with the number of cores you wish to use.

Distributed Computing

Quickstart

For those not wanting to read or understand the parallelism method, here are the basic required steps:

1. on the machine with the data, make a scratch directory on a local hard drive, cd to it, and run e2parallel.py dcserver --port=9990 --verbose=2

2. make another scratch directory on a local hard drive, cd to it, and run e2parallel.py dcclient --host=<server hostname>

3. repeat #2 for each core or machine you want to run tasks on
4. run your parallel job, like 'e2refine.py' with the --parallel=dc:localhost:9990

Notes

• If you need to kill the server and restart it for some reason, that's fine. As long as it is restarted within about 5 minutes, it should be harmless
• Make sure the same version of EMAN2 on all machines, if multiple machines are being used as clients
• If you need to stop the 'e2refine' program, you can run 'e2parallel.py killall' to cancel any pending jobs on the server after stopping e2refine.
• You can add or remove clients at any time during a run
• When you are done running jobs, kill the server, then run 'e2parallel.py dckillclients' from the server directory, and let it run for a minute or two. This will tell the clients to shut down. If you plan to do another run relatively soon, you can just leave the server and clients running.

You should really consider reading the detailed instructions below, though :^)

Introduction

This is the sort of parallelism made famous by projects like SETI-at-home and Folding-at-Home. The general idea is that you have a list of small jobs to do, and a bunch of computers with spare cycles willing to help out with the computation. The number of computers willing to do computations may vary with time, and possibly may agree to do a computation, but then fail to complete it. This is a very flexible parallelism model, which can be adapted to both individual computers with multiple cores as well as linux clusters or sets of workstations laying around the lab.

There are 3 components to this system:

User Application (customer) <==> Server <==> Compute Nodes (client)

The user application (e2refine.py for example) builds a list of computational tasks that it needs to have completed, then sends the list to the server. Compute nodes with nothing to do then contact the server and request tasks to compute. The server sends the tasks out to the clients. When the client finishes the requested computation, results are sent back to the server. The user application then requests the results from the server and completes processing. As long as the number of tasks to complete is larger than the number of clients servicing requests, this is an extremely efficient infrastructure.

Internally things are somewhat more complicated and tackle issues such as data caching on the clients, how to handle clients that die in the middle of processing, etc., but the basic concept is quite straightforward.

With any of the e2parallel.py commands below, you may consider adding the --verbose=1 option to see more of what it's doing.

How to use Distributed Computing in EMAN2

To use distributed computing, there are three basic steps:

• Run a server on a machine that the clients can communicate with
• Run some number of clients pointing at the server
• run an EMAN2 program with the --parallel option

What follows are specific instructions for doing this under 3 different scenarios.

Using DC on a single multi-core workstation

• Ideally your data will be stored on a hard drive physically connected to the workstation (not on a shared network drive)
• make an empty directory on a local hard drive

• Run a server on the workstation e2parallel.py dcserver from the empty directory you just created

• The server will print a message saying what port it's running on. This will usually be 9990. If it is something else, make a note of it.

• Run one client for each core you want to use for processing : e2parallel.py dcclient --server=localhost --port=9990 (replace the port with the correct number if necessary)

• Run your EMAN2 programs with the option --parallel=dc:localhost:9990 (again, use the right port number)

Using DC on a linux cluster

• The server should run on the node (often the head node or a specialized 'storage node') with a direct physical connection to the storage
• If you want to use clients from multiple clusters, then remember all of the clients must be able to make a network connection to the server machine

• Run a server on the head-node e2parallel.py dcserver in an empty directory on the local hard drive

• The server will print a message saying what port it's running on. This will usually be 9990. If it is something else, make a note of it.

• Run one client for each core you want to use for processing on each node : e2parallel.py dcclient --server=<server> --port=9990 (replace the server hostname and port with the correct values)

• Run your EMAN2 programs with the option --parallel=dc:<server>:9990 (again, use the right port number and server hostname)

Using DC on a set of workstations

• The server should run on a computer with a direct physical connection to the storage
• All of the clients must be able to make a network connection to the server machine

• Run a server on the desired machine e2parallel.py dcserver in an empty directory on the local hard drive

• The server will print a message saying what port it's running on. This will usually be 9990. If it is something else, make a note of it.

• Run one client for each core you want to use for processing on each computer : e2parallel.py dcclient --server=<server> --port=9990 (replace the server hostname and port with the correct values)

• Run your EMAN2 programs with the option --parallel=dc:<server>:9990 (again, use the right port number and server hostname)

For all of the above, once you have finished running your jobs, kill the server, then run 'e2parallel.py dckillclients' from the same directory. When it stops spewing out 'client killed' messages, you can kill this server.

IF THIS IS NOT WORKING FOR YOU, PLEASE FOLLOW THESE DEBUGGING INSTRUCTIONS

MPI

Sorry, we haven't had a chance to finish this yet. For the moment you will have to use the Distributed Computing mode on clusters.