we have a single threaded application that simulates the interaction of a hundred of thousands of objects over time with the shared memory model.
obviously, it suffers from its inability to scale over multi CPU hardware.
after reading a little about agent based modeling and functional programming/actor model I was considering a rewrite with the message-passing paradigm.
the idea is very simple - each object will be an actor and their interactions will be messages so that the simulation could happen in parallel. given a configuration of objects at a certain time - its future consequences can be easily computed.
the question is how to model the time:
for example let's assume the the behavior of object X depends on A and B, as the actors and the messages calculations order is not guaranteed it could be that when X is to be computed A has already sent its message to X but B didn't.
how to make sure the computation happens correctly?
I hope the question is clear
thanks in advance.
Your approach of using message passing to parallelize a (discrete-event?) simulation is well-known and does not require a functional style per se (although, of course, this does not prevent you to implement it like that).
The basic problem you describe w.r.t. to the timing of events is also known as the local causality constraint (see, for example, this textbook). Basically, you need to use a synchronization protocol to ensure that each object (or agent) processes its messages in the right order. In the domain of parallel discrete-event simulation, such objects are called logical processes, and they communicate via events (i.e. time-stamped messages).
Correctly implementing a synchronization protocol for these events is challenging and the right choice of protocol is highly application-specific. For example, one important factor is the average amount of computation required per event: if there is little computation required, the communication costs dominate the overall execution time and it will be hard to scale the simulation.
I would therefore recommend to look for existing solutions/libraries on top of the actors framework you intend to use before starting from scratch.
Related
I realize that it will be a basic concept, but it would be helpful if anyone could explain if fast data access related to the availability (A) in CAP theorem. Fast data access is an important feature expected of Big Data systems. And does the various K-Access and K-grouping method all a part of it.
Availability in CAP theorem is about whether you can access your data even if there is failure in the hardware (e.g. network outage, node outage etc).
Fast access to large volume of data is important feature observed in most of the big data systems. But, it should not confused with availability as described above.
Availability in CAP theorem means that all your requests will receive a response, but does not specify when or how accurate. Nor does it specify what fast could mean.
Every request receives a (non-error) response – without guarantee that
it contains the most recent write
Keep in mind that this theorem enforces strict guarantees. For example, systems can guarantee C and A, and still be good at P most of the time.
I am working on my bachelor's final project, which is about the comparison between Apache Spark Streaming and Apache Flink (only streaming) and I have just arrived to "Physical partitioning" in Flink's documentation. The matter is that in this documentation it doesn't explain well how this two transformations work. Directly from the documentation:
shuffle(): Partitions elements randomly according to a uniform distribution.
rebalance(): Partitions elements round-robin, creating equal load per partition. Useful for performance optimisation in the presence of data skew.
Source: https://ci.apache.org/projects/flink/flink-docs-release-1.2/dev/datastream_api.html#physical-partitioning
Both are automatically done, so what I understand is that they both redistribute equally (shuffle() > uniform distribution & rebalance() > round-robin) and randomly the data. Then I deduce that rebalance() distributes the data in a better way ("equal load per partitions") so the tasks have to process the same amount of data, but shuffle() may create bigger and smaller partitions. Then, in which cases might you prefer to use shuffle() than rebalance()?
The only thing that comes to my mind is that probably rebalance()requires some processing time so in some cases it might use more time to do the rebalancing than the time it will improve in the future transformations.
I have been looking for this and nobody has talked about this, only in a mailing list of Flink, but they don't explain how shuffle() works.
Thanks to Sneftel who has helped me to improve my question asking me things to let me rethink about what I wanted to ask; and to Till who answered quite well my question. :D
As the documentation states, shuffle will randomly distribute the data whereas rebalance will distribute the data in a round robin fashion. The latter is more efficient since you don't have to compute a random number. Moreover, depending on the randomness, you might end up with some kind of not so uniform distribution.
On the other hand, rebalance will always start sending the first element to the first channel. Thus, if you have only few elements (fewer elements than subtasks), then only some of the subtasks will receive elements, because you always start to send the first element to the first subtask. In the streaming case this should eventually not matter because you usually have an unbounded input stream.
The actual reason why both methods exist is a historically reason. shuffle was introduced first. In order to make the batch an streaming API more similar, rebalance was then introduced.
This statement by Flink is misleading:
Useful for performance optimisation in the presence of data skew.
Since it's used to describe rebalance, but not shuffle, it suggests it's the distinguishing factor. My understanding of it was that if some items are slow to process and some fast, the partitioner will use the next free channel to send the item to. But this is not the case, compare the code for rebalance and shuffle. The rebalance just adds to next channel regardless how busy it is.
// rebalance
nextChannelToSendTo = (nextChannelToSendTo + 1) % numberOfChannels;
// shuffle
nextChannelToSendTo = random.nextInt(numberOfChannels);
The statement can be also understood differently: the "load" doesn't mean actual processing time, just the number of items. If your original partitioning has skew (vastly different number of items in partitions), the operation will assign items to partitions uniformly. However in this case it applies to both operations.
My conclusion: shuffle and rebalance do the same thing, but rebalance does it slightly more efficiently. But the difference is so small that it's unlikely that you'll notice it, java.util.Random can generate 70m random numbers in a single thread on my machine.
I'm new to MPI and I'm trying to understand how MPI (and specifically OpenMPI) work in order to reason about the performance of my system.
I've tried to find resources online to help me understand things a little better, but haven't had much luck. I thought I'd come here.
Right now my question is simple: if I have 3 nodes (1 master, 2 clients) and I issue an MPI_Gather, does the root process handle incoming data sequentially or concurrently? In other words, if processes 1 is the first to make a connection with processes 0, will process 2 have to wait until processes 1 is done sending its data before it can start to send its data?
Thanks!
There are multiple components in Open MPI that implement collective operations and some of them provide multiple algorithms for the implementation of each operation.
What you are most likely interested in is the tuned component of the coll framework as that is what Open MPI uses by default. tuned implements all collectives using point-to-point operations and provides several algorithms for gather:
linear with synchronisation - used when messages are large to mid-size
binomial - used when the number of processes is large or the message size is small
basic linear - used in all other cases
The performance of each algorithm depends strongly on the particular combination of message size and number of ranks, therefore the library comes with a set of heuristics that tries to determine the best algorithm based on the data size and the size of the communicator (as indicated above). There are several mechanisms to override the heuristics and either force a certain algorithm or provide a list of custom algorithm selection rules.
The basic linear algorithm simply has the root loop over all other ranks receiving their messages in sequence. In that case, rank 2 won't be able to send its chunk before rank 1 since the root will first receive the message from rank 1 and only then move on to rank 2.
The linear with synchronisation algorithm splits the chunks into two pieces each. The first pieces are collected in sequence just like in the basic linear algorithm. The second pieces are collected asynchronously using non-blocking receives.
The binomial algorithm arranges the ranks as a binomial tree. The processes at the nodes of the tree receive the chunks from the lower levels and aggregate them into larger chunks that then get passed to the upper levels until they reach the root rank.
You can find the source code of the tuned module in the ompi/mca/coll/tuned folder of the Open MPI source tree. In the development branch, part of the tuned component got promoted to the base implementation of the collective framework and the code for the gather is to be found in ompi/mca/coll/base instead.
Hristo's answer is of course excellent, but I would like to offer a different point of view.
Contrary to your expectation, the question is not simple. It isn't even possible to specifically answer it without knowing more system specifics, as Hristo pointed out. That doesn't mean the question is invalid, but you should start to reason about performance on a different level.
First, consider the complexity of a the gather operation: The total network transfer to the root as well as the memory requirements are linearly growing with the number of processes in the communicator. This naturally limits scalability.
Second, you may assume that your MPI implementation does implement MPI_Gather in the most efficient way possible - better than you could do it by hand. This assumption may very well be wrong, but it is the best starting point to write your program.
Now when you have your program, you should measure and see where time is spent - or wasted. For that you should an MPI performance analysis tools. Now if you have identified that your Gather has a significant impact on performance, you can go ahead and try to optimize that: But to do so, first consider if you can structure your communication conceptually better, e.g. by somehow removing the computation all together or using a clever reduction instead. If you still need to stick to the gather: go ahead and tune your MPI implementation. Afterwards verify that your optimization did indeed improve performance on your specific system.
Is it possible to estimate the heat generated by an individual process in runtime.
Temperature readings of the processor is easily accessible but what I need is process specific information.
Is it possible to map information such as cpu utilization, io, running time, memory usage etc to get some kind of an estimate?
I'm gonna say no. Because the overall temperature of your system components isn't a simple mathematical equation with everything that's moving and switching either.
Heat generated by and inside a computer is dependent on many external factors like hardware setup, ambient temperature of the room, possibly the age of the components, is there dust on them or in the fans, was the cooling paste correctly applied on the CPU or elsewhere, where heat sinks are present, how is heat being dissipated etc.etc.. In short, again, no.
Additionally, your computer runs a LOT of processes at any given time apart from the ones that you control (and "control" is a relative term). Even if it is possible to access certain sensory data for individual components (like you can see to some extent in the BIOS) then interpolating one single process' generated temperature in regard to the total is, well, impossible.
At the lowest levels (gate networks, control signalling etc.), an external individual no longer has any means to observe or measure what's going on but there as well, things are in a changing state, a variable amount of electricity is being used and thus a variable amount of heat generated.
Pertaining to your second question: that's basically what your task manager does. There are countless examples and articles on the internet on how to get that done in a plethora of programming languages.
That is, unless some of the actually smart people in this merry little community of keytappers and screengazers say that it IS actually possible, at which point I will be thoroughly amazed...
EDIT: Monitoring the processes is a first step in what you're looking for. take a look at How to detect a process start & end using c# in windows? and be sure to follow up on duplicates like the one mentioned by Hans.
You could take a look at PowerTOP or some other tool that monitors power usage. I am not sure how accurate it is across different systems but a power estimation should provide at least some relative information as the heat generated assuming the processes you are comparing are running in similar manners on hardware. In reality there are just too many factors to predict power, much less heat, effectively but you may be able to get an idea of the usage.
I have a filtering algorithm that needs to be applied recursively and I am not sure if MapReduce is suitable for this job. W/o giving too much away, I can say that each object that is being filtered is characterized by a collection if ordered list or queue.
The data is not huge, just about 250MB when I export from SQL to
CSV.
The mapping step is simple: the head of the list contains an object that can classify the list as belonging to one of N mapping nodes. the filtration algorithm at each node works on the collection of lists assigned to the node and at the end of the filtration, either a list remains the same as before the filtration or the head of the list is removed.
The reduce function is simple too: all the map jobs' lists are brought together and may have to be written back to disk.
When all the N nodes have returned their output, the mapping step is repeated with this new set of data.
Note: N can be as much as 2000 nodes.
Simple, but it requires perhaps up to a 1000 recursions before the algorithm's termination conditions are met.
My question is would this job be suitable for Hadoop? If not, what are my options?
The main strength of Hadoop is its ability to transparently distribute work on a large number of machines. In order to fully benefit from Hadoop your application has to be characterized, at least by the following three things:
work with large amounts of data (data which is distributed in the cluster of machines) - which would be impossible to store on one machine
be data-parallelizable (i.e. chunks of the original data can be manipulated independently from other chunks)
the problem which the application is trying to solve lends itself nicely to the MapReduce (scatter - gather) model.
It seems that out of these 3, your application has only the last 2 characteristics (with the observation that you are trying to recursively use a scatter - gather procedure - which means a large number of jobs - equal to the recursion depth; see last paragraph why this might not be appropriate for hadoop).
Given the amount of data you're trying to process, I don't see any reason why you wouldn't do it on a single machine, completely in memory. If you think you can benefit from processing that small amount of data in parallel, I would recommend focusing on multicore processing than on distributed data intensive processing. Of course, using the processing power of a networked cluster is tempting but this comes at a cost: mainly the time inefficiency given by the network communication (network being the most contended resource in a hadoop cluster) and by the I/O. In scenarios which are well-fitted to the Hadoop framework these inefficiency can be ignored because of the efficiency gained by distributing the data and the associated work on that data.
As I can see, you need 1000 jobs. The setup and the cleanup of all those jobs would be an unnecessary overhead for your scenario. Also, the overhead of network transfer is not necessary, in my opinion.
Recursive algos are hard in the distributed systems since they can lead to a quick starvation. Any middleware that would work for that needs to support distributed continuations, i.e. the ability to make a "recursive" call without holding the resources (like threads) of the calling side.
GridGain is one product that natively supports distributed continuations.
THe litmus test on distributed continuations: try to develop a naive fibonacci implementation in distributed context using recursive calls. Here's the GridGain's example that implements this using continuations.
Hope it helps.
Q&D, but I suggest you read a comparison of MongoDB and Hadoop:
http://www.osintegrators.com/whitepapers/MongoHadoopWP/index.html
Without knowing more, it's hard to tell. You might want to try both. Post your results if you do!