I am trying to understand how indexing can be optimized on elasticsearch. Let me clarify my needs;
I have two indices rigth now. Lets say, indexA and indexB ( Two indices can be seen approximately same size)
I have 6 machines dedicated to elasticsearch (we can say exactly the same hardware)
The most important part of my elasticsearch usage is on writing since I am doing heavy writing on real time.
So my question is, how I can I optimize the writing operation using those 6 machines ?
Should I separate machines into two part like 3 machines for indexA and 3 machines for indexB ?
or
Should I use all of 6 machines in order to index indexA and indexB ?
and
What else should I need to give attention in order to optimize write operations ?
Thank you in advance
It depends, but let me take to a direction as per your problem statement which led to following assumptions:
you want to do more write operations (not worried about search performance)
both the indices are in the same cluster
in future more systems can get added
For better indexing performance first thing is you may want to have single shard for your index (unless you are using routing). But since you have 6 servers having single shard will be waste of resources so you can assign 3 shard to each of indexA and indexB. This is for current scenario but it is recommended to have ~10 shards(for future scalibility and your data size dependent)
Turn off the replica (if possible as index requests wait for the replicas to respond before returning). Though in production environment it is highly recommended to have at least one replica for high availability.
Set refresh rate to "-1" or at least to a larger figure say "30m". (You will lose NRT search if you do so but as you have mentioned you are concerned about indexing)
Turn of index warmers if you have any.
avoid using "doc_values" for your field mapping. (though it is beneficial for reducing memory footprint during search time it will increase your index time as it prepares field values during indexing)
If possible/not required disable "norms" in your mapping
Lastly read this.
Word of caution: some of the approach above will impact your search performance.
Related
I'm rather inexperienced with databases and have just read about the "n+1 selects issue". My follow-up question: Assuming the database resides on the same machine as my program, is cached in RAM and properly indexed, why is the n+1 query pattern slow?
As an example let's take the code from the accepted answer:
SELECT * FROM Cars;
/* for each car */
SELECT * FROM Wheel WHERE CarId = ?
With my mental model of the database cache, each of the SELECT * FROM Wheel WHERE CarId = ? queries should need:
1 lookup to reach the "Wheel" table (one hashmap get())
1 lookup to reach the list of k wheels with the specified CarId (another hashmap get())
k lookups to get the wheel rows for each matching wheel (k pointer dereferenciations)
Even if we multiply that by a small constant factor for an additional overhead because of the internal memory structure, it still should be unnoticeably fast. Is the interprocess communication the bottleneck?
Edit: I just found this related article via Hacker News: Following a Select Statement Through Postgres Internals. - HN discussion thread.
Edit 2: To clarify, I do assume N to be large. A non-trivial overhead will add up to a noticeable delay then, yes. I am asking why the overhead is non-trivial in the first place, for the setting described above.
You are correct that avoiding n+1 selects is less important in the scenario you describe. If the database is on a remote machine, communication latencies of > 1ms are common, i.e. the cpu would spend millions of clock cycles waiting for the network.
If we are on the same machine, the communication delay is several orders of magnitude smaller, but synchronous communication with another process necessarily involves a context switch, which commonly costs > 0.01 ms (source), which is tens of thousands of clock cycles.
In addition, both the ORM tool and the database will have some overhead per query.
To conclude, avoiding n+1 selects is far less important if the database is local, but still matters if n is large.
Assuming the database resides on the same machine as my program
Never assume this. Thinking about special cases like this is never a good idea. It's quite likely that your data will grow, and you will need to put your database on another server. Or you will want redundancy, which involves (you guessed it) another server. Or for security, you might want not want your app server on the same box as the DB.
why is the n+1 query pattern slow?
You don't think it's slow because your mental model of performance is probably all wrong.
1) RAM is horribly slow. Your CPU is wasting around 200-400 CPU cycles each time it needs to read something something from RAM. CPUs have a lot of tricks to hide this (caches, pipelining, hyperthreading)
2) Reading from RAM is not "Random Access". It's like a hard drive: sequential reads are faster.
See this article about how accessing RAM in the right order is 76.6% faster http://lwn.net/Articles/255364/ (Read the whole article if you want to know how horrifyingly complex RAM actually is.)
CPU cache
In your "N+1 query" case, the "loop" for each N includes many megabytes of code (on client and server) swapping in and out of caches on each iteration, plus context switches (which usually dumps the caches anyway).
The "1 query" case probably involves a single tight loop on the server (finding and copying each row), then a single tight loop on the client (reading each row). If those loops are small enough, they can execute 10-100x faster running from cache.
RAM sequential access
The "1 query" case will read everything from the DB to one linear buffer, send it to the client who will read it linearly. No random accesses during data transfer.
The "N+1 query" case will be allocating and de-allocating RAM N times, which (for various reasons) may not be the same physical bit of RAM.
Various other reasons
The networking subsystem only needs to read one or two TCP headers, instead of N.
Your DB only needs to parse one query instead of N.
When you throw in multi-users, the "locality/sequential access" gets even more fragmented in the N+1 case, but stays pretty good in the 1-query case.
Lots of other tricks that the CPU uses (e.g. branch prediction) work better with tight loops.
See: http://blogs.msdn.com/b/oldnewthing/archive/2014/06/13/10533875.aspx
Having the database on a local machine reduces the problem; however, most applications and databases will be on different machines, where each round trip takes at least a couple of milliseconds.
A database will also need a lot of locking and latching checks for each individual query. Context switches have already been mentioned by meriton. If you don't use a surrounding transaction, it also has to build implicit transactions for each query. Some query parsing overhead is still there, even with a parameterized, prepared query or one remembered by string equality (with parameters).
If the database gets filled up, query times may increase, compared to an almost empty database in the beginning.
If your database is to be used by other application, you will likely hammer it: even if your application works, others may slow down or even get an increasing number of failures, such as timeouts and deadlocks.
Also, consider having more than two levels of data. Imagine three levels: Blogs, Entries, Comments, with 100 blogs, each with 10 entries and 10 comments on each entry (for average). That's a SELECT 1+N+(NxM) situation. It will require 100 queries to retrieve the blog entries, and another 1000 to get all comments. Some more complex data, and you'll run into the 10000s or even 100000s.
Of course, bad programming may work in some cases and to some extent. If the database will always be on the same machine, nobody else uses it and the number of cars is never much more than 100, even a very sub-optimal program might be sufficient. But beware of the day any of these preconditions changes: refactoring the whole thing will take you much more time than doing it correctly in the beginning. And likely, you'll try some other workarounds first: a few more IF clauses, memory cache and the like, which help in the beginning, but mess up your code even more. In the end, you may be stuck in a "never touch a running system" position, where the system performance is becoming less and less acceptable, but refactoring is too risky and far more complex than changing correct code.
Also, a good ORM offers you ways around N+1: (N)Hibernate, for example, allows you to specify a batch-size (merging many SELECT * FROM Wheels WHERE CarId=? queries into one SELECT * FROM Wheels WHERE CarId IN (?, ?, ..., ?) ) or use a subselect (like: SELECT * FROM Wheels WHERE CarId IN (SELECT Id FROM Cars)).
The most simple option to avoid N+1 is a join, with the disadvantage that each car row is multiplied by the number of wheels, and multiple child/grandchild items likely ending up in a huge cartesian product of join results.
There is still overhead, even if the database is on the same machine, cached in RAM and properly indexed. The size of this overhead will depend on what DBMS you're using, the machine it's running on, the amount of users, the configuration of the DBMS (isolation level, ...) and so on.
When retrieving N rows, you can choose to pay this cost once or N times. Even a small cost can become noticeable if N is large enough.
One day someone might want to put the database on a separate machine or to use a different dbms. This happens frequently in the business world (to be compliant with some ISO standard, to reduce costs, to change vendors, ...)
So, sometimes it's good to plan for situations where the database isn't lightning fast.
All of this depends very much on what the software is for. Avoiding the "select n+1 problem" isn't always necessary, it's just a rule of thumb, to avoid a commonly encountered pitfall.
We currently have an SQL Federated DB split over 10 shards in roughly equal portions of data, filtered by a Client ID.
At the moment we are experiencing performance problems executing filtered queries, for example, running a query for a specific Client can take over 3 minutes to return 4000 rows in some shards. However, running exactly same query in an unfiltered connection on the same shard returns within a timely 4 seconds. The one noticable aspect is that the shards experiencing the slow down tend to contain more Clients albeit with less data. The most likely performance inhibitor (I believe) would be indexing and something that ties into the Filtered / Unfiltered connection.
Having a search around I haven't found much information on query performance across shards / specific Indexing strategies on shards (apart from Azure apparently doesn't support Indexed Views). My impression (and hence need for clarification) is that Indexes are applied to all members of a shard and not on a member by member basis.
If the former then we're in a bit of a pickle, apart from resharding this particular shard which doesn't make sense considering the only difference is the number of clients, not the size of the data. A couple of things we're about to try are explicitly adding the filter to the Indexes or even adding the filter to each query. Safe to say, we're not happy moving away from a Filtered connection.
Has anyone else experienced this problem or could possibly provide some direction that an unfiltered connection significantly outperforms a filtered connection?
Thanks in advance...
Indexes in federations are applied on a Federation Member-by-Member basis. If you started with a single indexed member and performed a SPLIT operation, then the indexes are automatically applied to the products of the SPLIT. But if you have applied indexes after multiple members were created, you need to explicitly add indexes to each member.
So hopefully you are not in a pickle.
You probably want to consider alternatives to federations moving forward since the feature is not supported by the new SKUs announced in April.
I am seeking a way to find bottlenecks in SQL server and it seems that more than 32GB ram and more than 32 spindels on 8 cores are not enough. Are there any metrics, best practices or HW comparations (i.e. transactions per sec)? Our daily closure takes hours and I want it in minutes or realtime if possible. I was not able to merge more than 12k rows/sec. For now, I had to split the traffic to more than one server, but is it a proper solution for ~50GB database?
Merge is enclosed in SP and keeped as simple as it can be - deduplicate input, insert new rows, update existing rows. I found that the more rows we put into single merge the more rows per sec we get. Application server runs in more threads, and uses all the memory and processor on its dedicated server.
Follow a methodology like Waits and Queues to identify the bottlenecks. That's exactly what is designed for. Once you identified the bottleneck, you can also judge whether is a hardware provisioning and calibration issue (and if so, which hardware is the bottleneck), or if is something else.
The basic idea is to avoid having to do random access to a disk, both reading and writing. Without doing any analysis, a 50 GB database needs at least 50GB of ram. Then you have to make sure indexes are on a separate spindle from the data and the transaction logs, you write as late as possible, and critical tables are split over multiple spindles. Are you doing all that?
Im new to database design and need some guidance.
A lot of new data is inserted to my database throughout the day. (100k rows per day)
The data is never modified or deleted once it has been inserted.
How can I optimize this database for retrieval speed?
My ideas
Create two databases (and possible on different hard drives) and merge the two at night when traffic is low
Create some special indexes...
Your recommendation is highly appreciated.
UPDATE:
My database only has a single table.
100k/day is actually fairly low. 3M/month, 40M/year. You can store 10 years archive and not reach 1B rows.
The most important thing to choose in your design will be the clustered key(s). You need to make sure that they are narrow and can serve all the queries your application will normally use. Any query that will end up in table scan will completely trash your memory by fetching in the entire table. So, no surprises there, your driving factor in your design is the actual load you'll have: exactly what queries will you be running.
A common problem (more often neglected than not) with any high insert rate is that eventually every row inserted will have to be deleted. Not acknowledging this is a pipe dream. The proper strategy depends on many factors, but probably the best bet is on a sliding window partitioning scheme. See How to Implement an Automatic Sliding Window in a Partitioned Table. This cannot be some afterthought, the choice for how to remove data will permeate every aspect of your design and you better start making a strategy now.
The best tip I can give which all big sites use to speed up there website is:
CACHE CACHE CACHE
use redis/memcached to cache your data! Because memory is (blazingly)fast and disc I/O is expensive.
Queue writes
Also for extra performance you could queue up the writes in memory for a little while before flushing them to disc -> writting them to SQL database. Off course then you have the risk off losing data if you keep it in memory and your computer crashes or has power failure or something
Context missing
Also I don't think you gave us much context!
What I think is missing is:
architecture.
What kind of server are you having VPS/shared hosting.
What kind of Operating system does it have linux/windows/macosx
computer specifics like how much memory available, cpu etc.
a find your definition of data a bit vague. Could you not attach a diagram or something which explains your domain a little bit. For example something like
this using http://yuml.me/
Your requirements are way to general. For MS SQL server 100k (more or less "normal") records per days should not be a problem, if you have decent hardware. Obviously you want to write fast to the database, but you ask for optimization for retrieval performance. That does not match very well! ;-) Tuning a database is a special skill on its own. So you will never get the general answer you would like to have.
Keep in mind that I am a rookie in the world of sql/databases.
I am inserting/updating thousands of objects every second. Those objects are actively being queried for at multiple second intervals.
What are some basic things I should do to performance tune my (postgres) database?
It's a broad topic, so here's lots of stuff for you to read up on.
EXPLAIN and EXPLAIN ANALYZE is extremely useful for understanding what's going on in your db-engine
Make sure relevant columns are indexed
Make sure irrelevant columns are not indexed (insert/update-performance can go down the drain if too many indexes must be updated)
Make sure your postgres.conf is tuned properly
Know what work_mem is, and how it affects your queries (mostly useful for larger queries)
Make sure your database is properly normalized
VACUUM for clearing out old data
ANALYZE for updating statistics (statistics target for amount of statistics)
Persistent connections (you could use a connection manager like pgpool or pgbouncer)
Understand how queries are constructed (joins, sub-selects, cursors)
Caching of data (i.e. memcached) is an option
And when you've exhausted those options: add more memory, faster disk-subsystem etc. Hardware matters, especially on larger datasets.
And of course, read all the other threads on postgres/databases. :)
First and foremost, read the official manual's Performance Tips.
Running EXPLAIN on all your queries and understanding its output will let you know if your queries are as fast as they could be, and if you should be adding indexes.
Once you've done that, I'd suggest reading over the Server Configuration part of the manual. There are many options which can be fine-tuned to further enhance performance. Make sure to understand the options you're setting though, since they could just as easily hinder performance if they're set incorrectly.
Remember that every time you change a query or an option, test and benchmark so that you know the effects of each change.
Actually there are some simple rules which will get you in most cases enough performance:
Indices are the first part. Primary keys are automatically indexed. I recommend to put indices on all foreign keys. Further put indices on all columns which are frequently queried, if there are heavily used queries on a table where more than one column is queried, put an index on those columns together.
Memory settings in your postgresql installation. Set following parameters higher:
.
shared_buffers, work_mem, maintenance_work_mem, temp_buffers
If it is a dedicated database machine you can easily set the first 3 of these to half the ram (just be carefull under linux with shared buffers, maybe you have to adjust the shmmax parameter), in any other cases it depends on how much ram you would like to give to postgresql.
http://www.postgresql.org/docs/8.3/interactive/runtime-config-resource.html
http://wiki.postgresql.org/wiki/Performance_Optimization
The absolute minimum I'll recommend is the EXPLAIN ANALYZE command. It will show a breakdown of subqueries, joins, et al., all the time showing the actual amount of time consumed in the operation. It will also alert you to sequential scans and other nasty trouble.
It is the best way to start.
Put fsync = off in your posgresql.conf, if you trust your filesystem, otherwise each postgresql operation will be imediately written to the disk (with fsync system call).
We have this option turned off on many production servers since quite 10 years, and we never had data corruptions.