I am performing stress testing on SQL Server 2008 with JMeter.
I wish to improve a stored procedure that has to serve 20 requests per second.
The procedure takes an xml parameter and returns an xml result.
Should I use only one parameter value or test multiple scenarios?
My main doubts are:
recompilations of the procedure execution plan (this may slow down the procedure)
extraction of data from disk (not all necessary data may be hold in the main memory)
Designing a realistic Stress Test/Load Test in SQL Server is an art.
There are many factors that can impact performance:
Hardware: You need to run your tests against the the same hardware that you have defined your target (20 call per second). This includes disk configuration, redundancy, clustering, ... This is not always possible so you need to make it as close as possible however the more different your test environment becomes, the more unrealistic results can be. This means, for example, if you use 2 CPUs instead of 4, you cannot adjust the parameters accordingly.
Data load: in terms of number of the records you need to test, it is ideal to have around 30%-40% more of the maximum rows you expect in the tables.
Data and index distribution: It is a common mistake to load the server with a preset or completely random data. Both are wrong. The distribution of the values need to be realistic. For example distribution of the marital status is not the same across all possible values so you need to design your data generation to include this.
Index fragmentation: this is a tough one. Normally indexes are rebuilt overnight, but during the course of the day, indexes become fragmented so the performance can be very different during those times.
Concurrent load: A server could provide you with 20 requests per second, if it is the only call you are making to the database but as soon as you start making other calls, it all falls to pieces. The load need to include other related parts of the system.
Operation Load: It is absolutely no point to make 20 calls per second if the requests are all the same. You need to use Data Generation techniques to make the requests realistic not purely random.
If you are using C#, I have done this tool a while back which might help you with creating realistic random data.
Related
I have some performance tests for an index structure on some data. I will be comparing 2 indexes side-by-side (still not decided if I will be using 2 VMs). I require results to be as neutral as possible of course, so I have these kinds of questions which I would appreciate any input about... How can I ensure/control what is influencing the test? For example, caching effects/order of arrival from one test to another will influence the result. How can I measure these influences? How do I create a suitable warm-up? Or what kind of statistical techniques can I use to nullify such influences (I don't think just averages is enough)?
Before you start:
Make sure your tables and indices have just been freshly created and populated. This avoids issues with regard to fragmentation. Otherwise, if the data in one test is heavily fragmented, and the other is not, you might not be comparing apples to apples.
Make sure your tables are properly ANALYZEd. This makes sure that the query planner has proper statistics in all cases.
If you just want a comparison, and not a test under realistic use, I'd just do:
Cold-start your (virtual) machine. Wait a reasonable but fixed time (let's say 5 min, or whatever is reasonable for your system) so that all startup processes have taken place and do not interfere with the DB execution.
Perform test with index1, and measure time (this is timing where you don't have anything cached by either the database nor the OS).
If you're interested in results when there are cache effects: Perform test again 10 times (or any number of times as big as reasonable). Measure each time, to account for variability due to other processes running on the VM, and other contingencies.
Reboot your machine, and repeat the whole process for test2. There are methods to clean the OS cache; but they're very system dependent, and you don't have a way to clean the database cache. Check See and clear Postgres caches/buffers?.
If you are really (or mostly) interested in performance when there are no cache effects, you should perform the whole process several times. It's slow and tedious. If you're only interested in the case where there's (most probably) a cache effect, you don't need to restart again.
Perform an ANOVA (or any other statistical hypothesis test you might think more suited) to decide if your average time is statistically different or not.
You can see an example of performing several tests in the answer to a question about NOT NULL versus CHECK(xx NOT NULL).
As neutral as possible, then create two databases on the same instance of your database management system, then create the same tablespaces with data, using indexes on one instance but not the other.
The challenge with a VM is you have arbitrated access to your disk resources ( unless you have each VM pinned to a specific interface and disk set ). Because of this, your arbitration model could vary from one test to the next. The most neutral course, which removes the arbitration, is on physical hardware....and the same hardware in both cases.
I am trying to generate a table that holds a client age analysis at a certain time. My source data is from Pastel Evolution accounting system.
They have a table valued function [_efnAgedPostARBalancesSum] that takes 2 Parameters (date and client link) and returns Age1, Age2, etc for entered client link. I need to get the ageing for all the clients in the client table.
I managed to get it working by using cross apply as per below, but it takes a long time to execute. If I run the age analysis from within Pastel it takes about 20 seconds, in Sql it takes about 6 minutes.
The function is encrypted so I cannot see what it does. I am using SQL Server 2008 R2.
Is there a more efficient alternative to cross apply?
SELECT
f.AccountLink,
f.AccountBalance,
f.Age1,
f.Age2,
f.Age3,
f.Age4,
f.Age5,
f.Age6,
f.Age7
FROM
Client
CROSS APPLY [_efnAgedPostARBalancesSum] ('2014-09-30', Client.DCLink) AS f
It looks like an AR aging bucket function from the outside - and probably has custom bucket sizes (given the generic age1, age2, etc). They're notoriously compute intensive. Its the kind of query that often spawns the need for a separate BI database as an OLTP system is not ideal for analytical queries. Its not only slow to run, its also likely to be impacting other work in your OLTP system while this function is banging on it.
You can bet it's looking at the due dates from the various documents that contain balances due (very likely several sources). They might not all be indexed on the due date columns. Look for that first. If you run the query in SSMS with show plan on, it may suggest one or more indexes to speed the execution - right-click in the query window and select "Show Actual Query Plan". From this, you can at least discover the tables that are being touched and the predicates involved in gathering the data...and you might get lucky with indexing.
There's no telling how efficiently the function computes the buckets. If they're not using some kind of window functions, it can be kind of horrible. You might find it advantageous to write your own UDF to get only what you want. Since it's generic, it may have a lot more work to do to cover all the possible bases - something your organization may not need.
If it is an inline function, you might get some relief by asking only for the columns you really need to look at. They're returning (at least) 7 buckets and a lot of AR reporting and analysis needs only 3 (the 30, 60, 90 day buckets, for example). It might also be worth doing a little pre-analysis to find out what clients you need to apply the function to - to keep from having to run it against your whole client domain.
Just looking at the function name - makes me think it's not a documented API per se. Encryption reinforces this hunch. Not sure how badly you really want to use such a function - no telling how it might get refactored (or removed) going forward.
I am trying to perform some data processing in a GAE application over data that is stored in the Datastore. The bottleneck point is the throughput in which the query returns entities and I wonder how to improve the query's performance.
What I do in general:
everything works in a task queue, so we have plenty of time (10 minute deadline).
I run a query over the ndb entities in order to select which entities need to be processed.
as the query returns results, I group entities in batches of, say, 1000 and send them to another task queue for further processing.
the stored data is going to be large (say 500K-1M entities) and there is a chance that the 10 minutes deadline is not enough. Therefore, when the task is reaching the taskqueue deadline, I spawn a new task. This means I need an ndb.Cursor in order to continue the query from where it stopped.
The problem is the rate in which the query returns entities. I have tried several approaches and observed the following performance (which is too slow for my app):
Use fetch_page() in a while loop.
The code is straightforward
while has_more and theres_more_time:
entities, cursor, more = query.fetch_page(1000, ...)
send_to_process_queue(entities)
has_more = more and cursor
With this approach, it takes 25-30 seconds to process 10K entities. Roughly speaking, that is 20K entities per minute. I tried changing the page size or the class of the frontend instance; neither made any difference in performance.
Segment the data and fire multiple fetch_page_async() in parallel.
This approach is taken from here (approach C)
The overall performance remains the same as above. I tried with various number of segments (from 2 to 10) in order to have 2-10 parallel fetch_async() calls. In all cases, the overall time remained the same. The more parallel fetch_page_async() are called, the longer it takes for each one to complete. I also tried with 20 parallel fetches and it got worse. Changing the page size or the fronted instance class did not have and impact either.
Fetch everything with a single fetch() call.
Now this is the least suitable approach (if not unsuitable at all) as the instance may run out of memory, plus I don't get a cursor in case I need to spawn to another task (in fact I won't even have the ability to do so, the task will simply exceed the deadline). I tried this out of curiosity in order to see how it performs and I observed the best performance! It took 8-10 seconds for 10K entities, which is roughly be 60K entities per minute. Now that is approx. 3 times faster than fetch_page(). I wonder why this happens.
Use query.iter() in a single loop.
This is match like the first approach. This will make use of the query iterator's underlying generator, plus I can obtain a cursor from the iterator in case I need to spawn a new task, so it suits me. With the query iterator, it fetched 10K entities in 16-18 seconds, which is approx. 36-40K entities per minute. The iterator is 30% faster than fetch_page, but much slower that fetch().
For all the above approaches, I tried F1 and F4 frontend instances without any difference in Datastore performance. I also tried to change the batch_size parameter in the queries, still without any change.
A first question is why do fetch(), fetch_page() and iter() behave so differently and how to make either fetch_page() or iter() do equally well as fetch()? And then another critical question is whether these throughputs (20-60K entities per minute, depending on api call) are the best we can do in GAE.
I 'm aware of the MapReduce API but I think it doesn't suit me. AFAIK, the MapReduce API doesn't support queries and I don't want to scan all the Datastore entities (it's will be too costly and slow - the query may return only a few results). Last, but not least, I have to stick to GAE. Resorting to another platform is not an option for me. So the question really is how to optimize the ndb query.
Any suggestions?
In case anyone is interested, I was able to significantly increase the throughput of the data processing by re-designing the component - it was suggested that I change the data models but that was not possible.
First, I segmented the data and then processed each data segment in a separate taskqueue.Task instead of calling multiple fetch_page_async from a single task (as I described in the first post). Initially, these tasks were processed by GAE sequentially utilizing only a single Fx instance. To achieve parallelization of the tasks, I moved the component to a specific GAE module and used basic scaling, i.e. addressable Bx instances. When I enqueue the tasks for each data segment, I explicitly instruct which basic instance will handle each task by specifying the 'target' option.
With this design, I was able to process 20.000 entities in total within 4-5 seconds (instead of 40'-60'!), using 5 B4 instances.
Now, this has additional costs because of the Bx instances. We 'll have to fine tune the type and number of basic instances we need.
The new experimental Data Processing feature (an AppEngine API for MapReduce) might be suitable. It uses automatic sharding to execute multiple parallel worker processes, which may or may not help (like the Approach C in the other linked question).
Your comment about "no need to scan all entities" triggers the thought that custom indexes could help your queries. That may entail schema changes to store the data in a less normal form.
Design a solution from the output perspective - what the simplest query is that produces the required results, then what the entity structure is to support such a query, then what work is needed to create and maintain such an entity structure from the current data.
I'm working on a web application where the user provides parameters, and these are used to produce a list of the top 1000 items from a database of up to 20 million rows. I need all top 1000 items at once, and I need this ranking to happen more or less instantaneously from the perspective of the user.
Currently, I'm using a MySQL with a user-defined function to score and rank the data, then PHP takes it from there. Tested on a database of 1M rows, this takes about 8 seconds, but I need performance around 2 seconds, even for a database of up to 20M rows. Preferably, this number should be lower still, so that decent throughput is guaranteed for up to 50 simultaneous users.
I am open to any process with any software that can process this data as efficiently as possible, whether it is MySQL or not. Here are the features and constraints of the process:
The data for each row that is relevant to the scoring process is about 50 bytes per item.
Inserts and updates to the DB are negligible.
Each score is independent of the others, so scores can be computed in parallel.
Due to the large number of parameters and parameter values, the scores cannot be pre-computed.
The method should scale well for multiple simultaneous users
The fewer computing resources this requires, in terms of number of servers, the better.
Thanks
A feasible approach seems to be to load (and later update) all data into about 1GB RAM and perform the scoring and ranking outside MySQL in a language like C++. That should be faster than MySQL.
The scoring must be relatively simple for this approache because your requirements only leave a tenth of a microsecond per row for scoring and ranking without parallelization or optimization.
If you could post query you are having issue with can help.
Although here are some things.
Make sure you have indexes created on database.
Make sure to use optimized queries and using joins instead of inner queries.
Based on your criteria, the possibility of improving performance would depend on whether or not you can use the input criteria to pre-filter the number of rows for which you need to calculate scores. I.e. if one of the user-provided parameters automatically disqualifies a large fraction of the rows, then applying that filtering first would improve performance. If none of the parameters have that characteristic, then you may need either much more hardware or a database with higher performance.
I'd say for this sort of problem, if you've done all the obvious software optimizations (and we can't know that, since you haven't mentioned anything about your software approaches), you should try for some serious hardware optimization. Max out the memory on your SQL servers, and try to fit your tables into memory where possible. Use an SSD for your table / index storage, for speedy deserialization. If you're clustered, crank up the networking to the highest feasible network speeds.
We are building a new application in .net 3.5 with SQL server database. The database is fairly large having around 60 tables with loads on data. The .net application have functionality to bring data into this database from data entry and from third party systems.
After all the data is available in database the system have to do lots of calculation. The calculation logic is pretty complex. All the data required for calculations is in database and the output also needs to be stored in database. The data gathering will happen every week and the calculation needs to be done every week to generate required reports.
Due to above scenario I was thinking do all these calculations using Stored Procedure. The problem is we need data independence also and stored procedure will not be able to provide us that. But if I do all this in .net by query database all the time, I don't think it will be able to finish the work quickly.
For example, I need to query one table which will return me 2000 rows then for each row I need to query another table which will return me 300 results than for each row of this I need to query multiple tables (around 10) to get required data, do the calculation and store the output in another table.
Now my question should I go ahead with stored-procedure solution and forget about database independence since performance is important. I also think development time will be much less if we use stored procedure solution. If any of client want this solution on say oracle database (because they don't want to maintain another database) then we port the stored procedures to oracle database and maintain two versions for any future changes/enhancements. Similarly other clients may ask for other databases.
The 2000 rows which I mentioned above is of product skus. The 300 rows I mentioned is of different attributes which we want to calculate, e.g. handling cost, transport cost, etc. The 10 tables I mentioned have information about currency conversion, unit conversion, network, area, company, sell price, number sold per day, etc. The resulting table stores all the information as a star schema for analysis and reporting purpose. The goal is to get any minute information about the product so that one know what attribute of a product selling is costing us money and where we can do the improvement.
I wouldn't consider doing the data manipulation anywhere other than in the database.
most people try to work with database data using looping algorithms. if you need real speed, think of your data as a SET of rows and you can update thousands of rows within a single update. I have rewritten so many cursor loops written by novice programmers into single update statements where the execution time was massively improved.
you say:
I need to query one table which will
return me 2000 rows then for each row
I need to query another table which
will return me 300 results than for
each row of this I need to query
multiple tables (around 10) to get
required data
from your question it looks like you are not using joins, and you are already thinking in loops. even if you do intend to loop, it is much better to write a query to join in all data necessary then loop over it. remember update and insert statements can have massively complex queries driving them. include in CASE statements, derived tables, conditional joins (LEFT OUTER JOIN) and you can just about solve any problem in a single update/insert.
Well without any specific details of what data you have in these tables, just a back of the napkin calculation shows that you're talking about processing over 6 million rows of information in the example you provided (2,000 rows * 300 rows * (1 row * 10 tables)).
Are all of these rows distinct, or are the 10 tables lookup information that has a relatively low cardinality? In other words, would it be possible to make a program that has the information from the 10 lookup tables in memory, and then just process the 300 row result set in memory to perform the calculations?
Also, I would be concerned about scalability -- if you do this in a stored procedure, it is guaranteed to be a serial process limited by the speed of the single database server. If you have the possibility of multiple copies of a client program, each processing a chunk of the 2,000 initial record set, then you can perform some of the calculations in parallel perhaps speeding up your overall processing time, as well as making it scalable for when your initial record set is 10 times larger.
Programming things like calculation code tend to be easier and more maintainable in C#. Also, normally keeping processing on the SQL Server to a minimum is a good practice since the database is the hardest to scale.
Having said that, from your description it sounds like the stored procedure approach is the way to go. When calculation code is dependent on large volumes of data, it's going to be more expensive to move the data off server for calculation. So unless you have reasonable ways of optimizing the dependent data (such as caching lookup tables?) then you are most likely going to find it more painful then it's worth to not use a stored proc.
Stored procedures every time, but as KM said within those stored procedures keep those iterations to minimum that is to say use joins in your SQL, relational databases are soooooo good at joining.
Database scalibility will be a small issue especially as it sounds like you'd be performing these calcualtions in a batch process.
Database independence doesn't really exist except for the most trivial of CRUD applications so if your initial requirement is to get this all working with SQL Server then leverage the tools that the RDBMS provides (after all your client will have spent a great deal of money on it). If (and it's a big if) a subsequent client really really doesn't want to use SQL Server then you'll have to bite the bullet and code it up in another flavour of stored procedure. But then as you identifed: "if I do all this in .net by query database all the time, I don't think it will be able to finish the work quickly." you've defered the expense of doing it until if and when required.
I would consider doing this in SQL Server Integration Services (SSIS). I'd put the calculations into SSIS, but leave the queries as stored procedures. This would provide you database independence - SSIS can process data from any database with an ODBC connection - as well as high performance. Only the simple SELECT statements would be in stored procedures, and those are the parts of the SQL standard most likely to be identical across multiple database products (assuming you stick to standard forms of query).