We need to read and count different types of messages/run
some statistics on a 10 GB text file, e.g a FIX engine
log. We use Linux, 32-bit, 4 CPUs, Intel, coding in Perl but
the language doesn't really matter.
I have found some interesting tips in Tim Bray's
WideFinder project. However, we've found that using memory mapping
is inherently limited by the 32 bit architecture.
We tried using multiple processes, which seems to work
faster if we process the file in parallel using 4 processes
on 4 CPUs. Adding multi-threading slows it down, maybe
because of the cost of context switching. We tried changing
the size of thread pool, but that is still slower than
simple multi-process version.
The memory mapping part is not very stable, sometimes it
takes 80 sec and sometimes 7 sec on a 2 GB file, maybe from
page faults or something related to virtual memory usage.
Anyway, Mmap cannot scale beyond 4 GB on a 32 bit
architecture.
We tried Perl's IPC::Mmap and Sys::Mmap. Looked
into Map-Reduce as well, but the problem is really I/O
bound, the processing itself is sufficiently fast.
So we decided to try optimize the basic I/O by tuning
buffering size, type, etc.
Can anyone who is aware of an existing project where this
problem was efficiently solved in any language/platform
point to a useful link or suggest a direction?
Most of the time you will be I/O bound not CPU bound, thus just read this file through normal Perl I/O and process it in single thread. Unless you prove that you can do more I/O than your single CPU work, don't waste your time with anything more. Anyway, you should ask: Why on Earth is this in one huge file? Why on Earth don't they split it in a reasonable way when they generate it? It would be magnitude more worth work. Then you can put it in separate I/O channels and use more CPU's (if you don't use some sort of RAID 0 or NAS or ...).
Measure, don't assume. Don't forget to flush caches before each test. Remember that serialized I/O is a magnitude faster than random.
This all depends on what kind of preprocessing you can do and and when.
On some of systems we have, we gzip such large text files, reducing them to 1/5 to 1/7 of their original size. Part of what makes this possible is we don't need to process these files
until hours after they're created, and at creation time we don't really have any other load on the machines.
Processing them is done more or less in the fashion of zcat thosefiles | ourprocessing.(well it's done over unix sockets though with a custom made zcat). It trades cpu time for disk i/o time, and for our system that has been well worth it. There's ofcourse a lot of variables that can make this a very poor design for a particular system.
Perhaps you've already read this forum thread, but if not:
http://www.perlmonks.org/?node_id=512221
It describes using Perl to do it line-by-line, and the users seem to think Perl is quite capable of it.
Oh, is it possible to process the file from a RAID array? If you have several mirrored disks, then the read speed can be improved. Competition for disk resources may be what makes your multiple-threads attempt not work.
Best of luck.
I wish I knew more about the content of your file, but not knowing other than that it is text, this sounds like an excellent MapReduce kind of problem.
PS, the fastest read of any file is a linear read. cat file > /dev/null should be the speed that the file can be read.
Have you thought of streaming the file and filtering out to a secondary file any interesting results? (Repeat until you have a manageble size file).
Basically need to "Divide and conquer", if you have a network of computers, then copy the 10G file to as many client PCs as possible, get each client PC to read an offset of the file. For added bonus, get EACH pc to implement multi threading in addition to distributed reading.
Parse the file once, reading line by line. Put the results in a table in a decent database. Run as many queries as you wish. Feed the beast regularly with new incoming data.
Realize that manipulating a 10 Gb file, transferring it across the (even if local) network, exploring complicated solutions etc all take time.
I have a co-worker who sped up his FIX reading by going to 64-bit Linux. If it's something worthwhile, drop a little cash to get some fancier hardware.
hmmm, but what's wrong with the read() command in C? Usually has a 2GB limit,
so just call it 5 times in sequence. That should be fairly fast.
If you are I/O bound and your file is on a single disk, then there isn't much to do. A straightforward single-threaded linear scan across the whole file is the fastest way to get the data off of the disk. Using large buffer sizes might help a bit.
If you can convince the writer of the file to stripe it across multiple disks / machines, then you could think about multithreading the reader (one thread per read head, each thread reading the data from a single stripe).
Since you said platform and language doesn't matter...
If you want a stable performance that is as fast as the source medium allows for, the only way I am aware that this can be done on Windows is by overlapped non-OS-buffered aligned sequential reads. You can probably get to some GB/s with two or three buffers, beyond that, at some point you need a ring buffer (one writer, 1+ readers) to avoid any copying. The exact implementation depends on the driver/APIs. If there's any memory copying going on the thread (both in kernel and usermode) dealing with the IO, obviously the larger buffer is to copy, the more time is wasted on that rather than doing the IO. So the optimal buffer size depends on the firmware and driver. On windows good values to try are multiples of 32 KB for disk IO. Windows file buffering, memory mapping and all that stuff adds overhead. Only good if doing either (or both) multiple reads of same data in random access manner. So for reading a large file sequentially a single time, you don't want the OS to buffer anything or do any memcpy's. If using C# there's also penalties for calling into the OS due to marshaling, so the interop code may need bit of optimization unless you use C++/CLI.
Some people prefer throwing hardware at problems but if you have more time than money, in some scenarios it's possible to optimize things to perform 100-1000x better on a single consumer level computer than a 1000 enterprise priced computers. The reason is that if the processing is also latency sensitive, going beyond using two cores is probably adding latency. This is why drivers can push gigabytes/s while enterprise software is ends stuck at megabytes/s by the time it's all done. Whatever reporting, business logic and such the enterprise software do can probably also be done at gigabytes/s on two core consumer CPU, if written like you were back in the 80's writing a game. The most famous example I've heard of approaching their entire business logic in this manner is the LMAX forex exchange, which published some of their ring buffer based code, which was said to be inspired by network card drivers.
Forgetting all the theory, if you are happy with < 1 GB/s, one possible starting point on Windows I've found is looking at readfile source from winimage, unless you want to dig into sdk/driver samples. It may need some source code fixes to calculate perf correctly at SSD speeds. Experiment with buffer sizes also.
The switches /h multi-threaded and /o overlapped (completion port) IO with optimal buffer size (try 32,64,128 KB etc) using no windows file buffering in my experience give best perf when reading from SSD (cold data) while simultaneously processing (use the /a for Adler processing as otherwise it's too CPU-bound).
I seem to recall a project in which we were reading big files, Our implementation used multithreading - basically n * worker_threads were starting at incrementing offsets of the file (0, chunk_size, 2xchunk_size, 3x chunk_size ... n-1x chunk_size) and was reading smaller chunks of information. I can't exactly recall our reasoning for this as someone else was desining the whole thing - the workers weren't the only thing to it, but that's roughly how we did it.
Hope it helps
Its not stated in the problem that sequence matters really or not. So,
divide the file into equal parts say 1GB each, and since you are using multiple CPUs, then multiple threads wont be a problem, so read each file using separate thread, and use RAM of capacity > 10 GB, then all your contents would be stored in RAM read by multiple threads.
Related
I am running some code in parallel by using a forking module in perl called Parallel::ForkManager. I have currently setting the maximum number of processes to 30:
my $pm = Parallel::ForkManager->new(30);
What would be an advisable maximum number of processes to create? I am doing this on a commercial grade Solaris server, but I still don't want to overload the system.
In downloading files, this really depends on
how many different hosts you're downloading from, and
how fast they will give you the requested files compared to your maximum bandwidth.
If you're downloading files from a single machine to a single machine on a local network, 2-3 is about max. If you're downloading files from 30 different servers on the internet, all of which are slow, but you have a fat pipe, then 30 might be reasonable.
There is no one universal right answer here. Unless you count "it depends."
The purpose of "downloading files" was mentioned, but in comments a while ago and I take the question as stated, to also be more general.
The only relevant measure is when you start reaching saturation in performance gains, with particular software on that system. The formal limits are huge and meaningless while rules of thumb are very general.
Let's imagine to run 10 processes and the time to complete the job drops 10 times. Increase to 20 processes and the time drops 20 times -- but for 30 processes the gain is the factor of 10. At this point we have loaded the system. Push further and the performance will degrade rapidly, and for everyone. At that point the server is overloaded, even though it allows, say, 1024 processes per user (and really ten or more times that for a server).
With a few processes per core the machine is engaged and I'd say that that is a good rule of thumb. However, it is too general. I doubt that you'd gain much in performance by going to that many processes, given the many other factors that affect it.
Accessing one web server The server's capability is the gospel. They may have posted how many requests per seconds they are happy with. Or they may have a limit on number of processes per user, say 10 or 20. If that means that many simultaneous downloads then that's your limit. But I'd be careful -- if the site is close and fast a request may complete in as little as 0.1 or 0.2 seconds. Then, with 10 processes you may be hitting the server 100 times a second. I do not recommend that. If there is no information I'd say keep it to a few requests per second. The performance and server load also depend on the content -- big downloads are different from pulling many skinny web pages. The I/O on your side may matter but I'd expect the server to set the limit. If you are going to use their service a lot why not send an email and ask what they are OK with.
I/O, network (many servers) or disk With network the performance depends on every piece of hardware in the path as well as on software. Nobody can tell without trying it out. The disk I/O is very complex. To add to trouble it is unclear whether it'd be your disks or network that is the bottleneck. I'd expect clear performance gains up to a few tens of processes, and probably fewer.
CPU or memory bound This may be easiest -- processing that can be broken up in parallel on 30 cores can enjoy close to a factor of 30 speedup (given no other bottlenecks). Going beyond the number of cores clearly leads to reduced performance gain. Concurrent (but not parallel) processing is far more complicated. If your code is memory intensive that is yet completely different.
Useful basic tools for assessing above components are iostat -xzn, netstat -I, and vmstat. But there is a bit of a curve to learning how to interpret their output and hopefully it doesn't come to that.
The conclusion is that you have to time it. Take your real application and time it running in one process. Do this 3 to 5 times and see the average (throw away obvious outliers). Then repeat with 5 processes, then with 10, etc. I'd expect that the trend will start slowing down far sooner than the 30 processors you mention. Once it gets to that the system is loaded and whoever works on it will notice. Very soon after that the performance will likely degrade rapidly. Proper benchmarking tools, like Benchmark, are far more sophisticated but this may well settle the issue. If you see strange or inconsistent behavior you may have to dig into details, starting with tools mentioned above.
What "overloaded" means is a bit unclear. I like to cap my use of resources well before other people are affected. But it may be possible to push it, in particular if you can run when it's quiet. I doubt that you'll keep having a worthy gain all the way to the number of available processors.
So there is no concern about "overloading" the server if you first time things. The performance limit will tell you when to stop. I'd say that your limit of 30 is very reasonable. Unless this is really about downloading files, in which case the web server is likely all that matters.
You should set the maximum number of processes to 60.
I was asked this question in a interview long time back in a design your own RTOS question. Is there a limitation to the number of processes a real time operating system can handle? What would cause this limitation? From what I know each process should have its own PC, call stack, heap, file descriptors, page tables, etc.. I assume the kernel has to keep track of the process using some data structure. Is the limitation derived from this data structure?
In most cases the amount of RAM available is the only limiting factor (as is the case in FreeRTOS), however in a few cases there are limitations imposed by the chosen scheduling algorithm. For example uCOS/II has (I think) a limitation of 255 because of the bitmap scheduler used - but even so 255 is more than you would ever need in a real time system of the type it is designed for.
I'm using MongoDB on a 32 bit production system, which sucks but it's out of my control right now. The challenge is to keep the memory usage under ~2.5GB since going over this will cause 32 bit systems to crash.
According to the mongoDB team, the best way to track the memory usage is to use your operating system's process tracking system (i.e. ps or htop on Unix systems; Process Explorer on Windows.) for virtual memory size.
The DB mainly consists of one table which is continually cycling data, i.e. receiving data at regular intervals from sensors, and every day a cron job wipes all data from before the last 3 days. Over a period of time, the memory usage slowly increases. I took some notes over time using db.serverStats(), db.lectura.totalSize() and ps, shown in the chart below. Note that the size of the table in question has reduced in the last month but the memory usage increased nonetheless.
Now, there is some scope for adjustment in how many days of data I store. Today I deleted basically half of the data, and then restarted mongodb, and yet the mem virtual / mem mapped and most importantly memory usage according to ps have hardly changed! Why do these not reduce when I wipe data (and restart)? I read some other questions where people said that mongo isn't really using all the memory that it might appear to be using, and that you can't clear the cache or limit memory use. But then how can I ensure I stay under the 2.5GB limit?
Unless there is a way to stem this dataset-size-irrespective gradual increase in memory usage, it seems to me that the 32-bit version of Mongo is unuseable. Note: I don't mind losing a bit of performance if it solves the problem.
To answer regarding why the mapped and virtual memory usage does not decrease with the deletes, the mapped number is actually what you get when you mmap() the entire set of data files. This does not shrink when you delete records, because although the space is freed up inside the data files, they are not themselves reduced in size - the files are just more empty afterwards.
Virtual will include journal files, and connections, and other non-data related memory usage also, but the same principle applies there. This, and more, is described here:
http://www.mongodb.org/display/DOCS/Checking+Server+Memory+Usage
So, the 2GB storage size limitation on 32-bit will actually apply to the data files whether or not there is data in them. To reclaim deleted space, you will have to run a repair. This is a blocking operation and will require the database to be offline/unavailable while it was run. It will also need up to 2x the original size in terms of free disk space to be able to run the repair, since it essentially represents writing out the files again from scratch.
This limitation, and the problems it causes, is why the 32-bit version should not be run in production, it is just not suitable. I would recommend getting onto a 64-bit version as soon as possible.
By the way, neither of these figures (mapped or virtual) actually represents your resident memory usage, which is what you really want to look at. The best way to do this over time is via MMS, which is the free monitoring service provided by 10gen - it will graph virtual, mapped and resident memory for you over time as well as plenty of other stats.
If you want an immediate view, run mongostat and check out the corresponding memory columns (res, mapped, virtual).
In general, when using 64-bit builds with essentially unlimited storage, the data will usually greatly exceed the available memory. Therefore, mongod will use all of the available memory it can in terms of resident memory (which is why you should always have swap configured to the OOM Killer does not come into play).
Once that is used, the OS does not stop allocating memory, it will just have the oldest items paged out to make room for the new data (LRU). In other words, the recycling of memory will be done for you, and the resident memory level will remain fairly constant.
Your options for stretching 32-bit are limited, but you can try some things. The thing that you run out of is address space, and the increases in the sizes of additional database files mean that you would like to avoid crossing over the boundary from "n" files to "n+1". It may be worth structuring your data into more or fewer databases so that you can get the maximum amount of actual data into memory and as little as possible "dead space".
For example, if your database named "mydatabase" consists of the files mydatabase.ns (the namespace file) at 16 MB, mydatabase.0 at 64 MB, mydatabase.1 at 128 MB and mydatabase.2 at 256 MB, then the next file created for this database will be mydatabase.3 at 512 MB. If instead of adding to mydatabase you instead created an additional database "mynewdatabase" it would start life with mynewdatabase.ns at 16 MB and mynewdatabase.0 at 64 MB ... quite a bit smaller than the 512 MB that adding to the original database would be. In fact, you could create 4 new databases for less space than would be consumed by adding a new file to the original database, and because the files are smaller they would be easier to fit into contiguous blocks of memory.
It is a well-known message that 32-bit should not be used for production.
Use 64-bit systems.
Point.
I have an Intel Core 2 Duo CPU and i was reading 3 files from my C: drive and showing
some matching values from the files onto a EditBox on Screen.The whole process takes 2 minutes.Then I thought of processing each file in a separate thread and then the whole process is taking 2.30 minutes !!! i.e 30 seconds more than single threaded processing.
I was expecting the other way around !I can see both the Graphs in CPU usage history.Some one please explain to me what is going on ?
here is my code snippet.
foreach (FileInfo file in FileList)
{
Thread t = new Thread(new ParameterizedThreadStart(ProcessFileData));
t.Start(file.FullName);
}
where processFileData is the method that process the files.
Thanks!
The root of the problem is that the files are on the same drive and, unlike your dual core processor, your hard drive can only do one thing at a time.
If you read two files simultaneously, the disk heads will jump from one file to the other and back again. Given that your hard drive can read each file in roughly 40 seconds, it now has the additional overhead of moving its disk head between the three separate files many times during the read.
The fastest way to read multiple files from a single hard drive is to do it all in one thread and read them one after another. This way, the head only moves once per file read (at the very beginning) and not multiple times per read.
To optimize this process, you'll either need to change your logic (do you really need to read the whole contents of all three files?). Or purchase a faster hard drive/put the 3 files in three different hard drives and use threading/use a raid.
If you read from disk using multiple threads, then the disk heads will bounce around from one part of the disk to another as each thread reads from a different part of the drive. That can reduce throughput significantly, as you've seen.
For that reason, it's actually often a better idea to have all disk accesses go through a single thread, to help minimize disk seeks.
If your task is I/O bound and if it needs to run often, you might look at a tool like "contig" to make sure the layout of your files on disk is optimized / contiguous.
If you processing is mostly IO bound and CPU bound it make sense it take same time or even more.
How do you compare those files ? You should think what is the bottleneck of you application? IO output/input, CPU, memory ...
The multithreading is only interesting for CPU bound processing. i.e. complex calculation, comparison of data in memory, sorting etc ...
Since your process is IO bound, you should let the OS do your threading for you. Look at FileStream.BeginRead() for an example how to queue up your reads. Your EndRead() method can spin up your next request to read your next block of data pointing to itself to handle each subsequent completed block.
Also, with you creating additional threads, the OS has to manage more threads. And if a different CPU happens to get picked to handle the completed read, you've lost all of the CPU caching where your thread originated.
As you've found, you can't "speed up" an application just by adding threads.
Could anyone explain to me the differences between multi-CPU, multi-core, and hyper-thread? I am always confused about these differences, and about the pros/cons of each architecture in different scenarios.
Here is my current understanding after learning online and learning from others' comments.
I think hyper-thread is the most inferior technology among them, but cheap. Its main idea is duplicate registers to save context switch time;
Multi processor is better than hyper-thread, but since different CPUs are on different chips, the communication between different CPUs is of longer latency than multi-core, and using multiple chips, there is more expense and more power consumption than with multi-core;
multi-core integrates all the CPUs on a single chip, so the latency of communication between different CPUs are greatly reduced compared with multi-processor. Since it uses one single chip to contain all CPUs, it consumer less power and is less expensive than a multi processor system.
Is this correct?
Multi-CPU was the first version: You'd have one or more mainboards with one or more CPU chips on them. The main problem here was that the CPUs would have to expose some of their internal data to the other CPU so they wouldn't get in their way.
The next step was hyper-threading. One chip on the mainboard but it had some parts twice internally so it could execute two instructions at the same time.
The current development is multi-core. It's basically the original idea (several complete CPUs) but in a single chip. The advantage: Chip designers can easily put the additional wires for the sync signals into the chip (instead of having to route them out on a pin, then over the crowded mainboard and up into a second chip).
Super computers today are multi-cpu, multi-core: They have lots of mainboards with usually 2-4 CPUs on them, each CPU is multi-core and each has its own RAM.
[EDIT] You got that pretty much right. Just a few minor points:
Hyper-threading keeps track of two contexts at once in a single core, exposing more parallelism to the out-of-order CPU core. This keeps the execution units fed with work, even when one thread is stalled on a cache miss, branch mispredict, or waiting for results from high-latency instructions. It's a way to get more total throughput without replicating much hardware, but if anything it slows down each thread individually. See this Q&A for more details, and an explanation of what was wrong with the previous wording of this paragraph.
The main problem with multi-CPU is that code running on them will eventually access the RAM. There are N CPUs but only one bus to access the RAM. So you must have some hardware which makes sure that a) each CPU gets a fair amount of RAM access, b) that accesses to the same part of the RAM don't cause problems and c) most importantly, that CPU 2 will be notified when CPU 1 writes to some memory address which CPU 2 has in its internal cache. If that doesn't happen, CPU 2 will happily use the cached value, oblivious to the fact that it is outdated
Just imagine you have tasks in a list and you want to spread them to all available CPUs. So CPU 1 will fetch the first element from the list and update the pointers. CPU 2 will do the same. For efficiency reasons, both CPUs will not only copy the few bytes into the cache but a whole "cache line" (whatever that may be). The assumption is that, when you read byte X, you'll soon read X+1, too.
Now both CPUs have a copy of the memory in their cache. CPU 1 will then fetch the next item from the list. Without cache sync, it won't have noticed that CPU 2 has changed the list, too, and it will start to work on the same item as CPU 2.
This is what effectively makes multi-CPU so complicated. Side effects of this can lead to a performance which is worse than what you'd get if the whole code ran only on a single CPU. The solution was multi-core: You can easily add as many wires as you need to synchronize the caches; you could even copy data from one cache to another (updating parts of a cache line without having to flush and reload it), etc. Or the cache logic could make sure that all CPUs get the same cache line when they access the same part of real RAM, simply blocking CPU 2 for a few nanoseconds until CPU 1 has made its changes.
[EDIT2] The main reason why multi-core is simpler than multi-cpu is that on a mainboard, you simply can't run all wires between the two chips which you'd need to make sync effective. Plus a signal only travels 30cm/ns tops (speed of light; in a wire, you usually have much less). And don't forget that, on a multi-layer mainboard, signals start to influence each other (crosstalk). We like to think that 0 is 0V and 1 is 5V but in reality, "0" is something between -0.5V (overdrive when dropping a line from 1->0) and .5V and "1" is anything above 0.8V.
If you have everything inside of a single chip, signals run much faster and you can have as many as you like (well, almost :). Also, signal crosstalk is much easier to control.
You can find some interesting articles about dual CPU, multi-core and hyper-threading on Intel's website or in a short article from Yale University.
I hope you find here all the information you need.
In a nutshell: multi-CPU or multi-processor system has several processors. A multi-core system is a multi-processor system with several processors on the same die. In hyperthreading, multiple threads can run on the same processor (that is the context-switch time between these multiple threads is very small).
Multi-processors have been there for 30 years now but mostly in labs. Multi-core is the new popular multi-processor. Server processors nowadays implement hyperthreading along with multi-processors.
The wikipedia articles on these topics are quite illustrative.
Hyperthreading is a cheaper and slower alternative to having multiple-cores
The Intel Manual Volume 3 System Programming Guide - 325384-056US September 2015 8.7 "INTEL HYPER-THREADING TECHNOLOGY ARCHITECTURE" describes HT briefly. It contains the following diagram:
TODO it is slower by how much percent in average in real applications?
Hyperthreading is possible because modern single CPUs cores already execute multiple instructions at once with the instruction pipeline https://en.wikipedia.org/wiki/Instruction_pipelining
The instruction pipeline is a separation of functions inside of a single core to ensure that each part of the circuit is used at any given time: reading memory, decoding instructions, executing instructions, etc.
Hyperthreading separates functions further by using:
a single backend, which actually runs the instructions with its pipeline.
Dual core has two backends, which explains the greater cost and performance.
two front-ends, which take two streams of instructions and order them in a way to maximize pipelining usage of the single backend by avoiding hazards.
Dual core would also have 2 front-ends, one for each backend.
There are edge cases where instruction reordering produces no benefit, making hyperthreading useless. But it produces a significant improvement in average.
Two hyperthreads in a single core share further cache levels (TODO how many? L1?) than two different cores, which share only L3, see:
Multiple threads and CPU cache
How are cache memories shared in multicore Intel CPUs?
The interface that each hyperthread exposes to the operating system is similar to that of an actual core, and both can be controlled separately. Thus cat /proc/cpuinfo shows me 4 processors, even though I only have 2 cores with 2 hyperthreads each.
Operating systems can however take advantage of knowing which hyperthreads are on the same core to run multiple threads of a given program on a single core, which might improve cache usage.
This LinusTechTips video contains a light-hearted non-technical explanation: https://www.youtube.com/watch?v=wnS50lJicXc
Multi-CPU is a bit like multicore, but communication can only happen through RAM, not L3 cache
This means that if possible, you want to partition tasks that use the same memory a lot for each separate CPU.
E.g. the following SBI-7228R-T2X blade server contains 4 CPUs, 2 on each node:
Source.
We can see that there seem to be 4 sockets for the CPUs, each covered by a heat sink, with one open.
I think across the nodes, they don't even share RAM memory and must communicate through some kind of networking, thus representing one further step up on the hyperthread/multicore/multi-CPU hierarchy, TODO confirm:
https://scicomp.stackexchange.com/questions/7530/difference-between-nodes-and-cpus-when-running-software-on-a-cluster
SLURM nodes, tasks, cores, and cpus
https://www.quora.com/In-High-Performance-Computing-what-exactly-is-the-difference-between-the-terms-%E2%80%9Ccores-%E2%80%9D-%E2%80%9Cprocessors-%E2%80%9D-%E2%80%9Cnodes-%E2%80%9D-and-%E2%80%9Cclusters%E2%80%9D