I was wondering if sharding is an alternate name for partial replication or not. What I have figured out that --
Partial Repl. – each data item has only copies at some but not all of the nodes (‘Sharding’?)
Pure Partial Repl. – has only copies of a subset of the data item but no node contains a full copy of the database
Hybrid Partial Repl. – a set of nodes are full replicas and another set of nodes are partial replicas
Partial replication is an interesting way, in which you distribute the data with replication from a master to slaves, each contains a portion of the data. Eventually you get an array of smaller DBs, read only, each contains a portion of the data. Reads can very well be distributed and parallelized.
But what about the writes?
Those are still clogged, in 1 big fat lazy master database, tasks as buffer management, locking, thread locks/semaphores, and recovery tasks - are the real bottleneck of the OLTP, they make writes impossible to scale... See more in my blog post here: http://database-scalability.blogspot.com/2012/08/scale-up-partitioning-scale-out.html. BTW - your topic right here just gave me a great idea for another post. I'll link to this question and give you the credit! :)
Sharding is where data appears only once, within an array of DBs. Each database is the complete owner of the data, data is read from there, data is written to there. This way, reads and writes are distributed and parallelized. Real scale-out can be acheived.
Sharding is a mess to handle, to maintain, it's hard as hell. ScaleBase (I work there), enable automatic transparent scale-out, just throw it in the middle and you'll have 10 DBs at the back, and it'll look like 1 to your app. Automatic, transparent super-sharding - in a box.
Sharding is a method of horizontal partitioning of a table. It doesn't related to replication.
Traditionally an RDBMS server located in the center of system with star like topology. That's why it becomes:
the single point of failure
the performance bottleneck of the system
To resolve issue #1 you use replication: if original server dies you fail over to a replica.
To resolve issue #2 you can:
use sharding
1.1 do sharding by yourself
1.2 use your RDBMS "out of the box" clustering mechanism
migrate to a NoSQL solution
Sharding allows you to scale out database to many servers by splitting the data among them. However sharding is a trade-off. It limits you in data joining/intersecting/etc.
You still have issue #1 if you use sharding. So it's a good practice to replicate sharded nodes.
Related
I am currently working on a website where, roughly 40 million documents and images should be served to it's users. I need suggestions on which method is the most suitable for storing content with subject to these requirements.
System should be highly available, scale-able and durable.
Files have to be stored permanently and users should be able to modify them.
Due to client restrictions, 3rd party object storage providers such as Amazon S3 and CDNs are not suitable.
File size of content can vary from 1 MB to 30 MB. (However about 90% of the files would be less than 2 MB)
Content retrieval latency is not much of a problem. Therefore indexing or caching is not very important.
I did some research and found out about the following solutions;
Storing content as BLOBs in databases.
Using GridFS to chunk and store content.
Storing content in a file server in directories using a hash and storing the metadata in a database.
Using a distributed file system such as GlusterFS or HDFS and storing the file metadata in a database.
The website is developed using PHP and Couchbase Community Edition is used as the database.
I would really appreciate any input.
Thank you.
I have been working on a similar system for last two years, the work is still in progress. However, requirements are slightly different from yours: modifications are not possible (I will try to explain why later), file sizes fall in range from several bytes to several megabytes, and, the most important one, the deduplication, which should be implemented both on the document and block levels. If two different users upload the same file to the storage, the only copy of the file should be kept. Also if two different files partially intersect with each other, it's necessary to store the only copy of the common part of these files.
But let's focus on your requirements, so deduplication is not the case. First of all, high availability implies replication. You'll have to store your file in several replicas (typically 2 or 3, but there are techniques to decrease data parity) on independent machines in order to stay alive in case if one of the storage servers in your backend dies. Also, taking into account the estimation of the data amount, it's clear that all your data just won't fit into a single server, so vertical scaling is not possible and you have to consider partitioning. Finally, you need to take into account concurrency control to avoid race conditions when two different clients are trying to write or update the same data simultaneously. This topic is close to the concept of transactions (I don't mean ACID literally, but something close). So, to summarize, these facts mean that you're are actually looking for distributed database designed to store BLOBs.
On of the biggest problems in distributed systems is difficulties with global state of the system. In brief, there are two approaches:
Choose leader that will communicate with other peers and maintain global state of the distributed system. This approach provides strong consistency and linearizability guarantees. The main disadvantage is that in this case leader becomes the single point of failure. If leader dies, either some observer must assign leader role to one of the replicas (common case for master-slave replication in RDBMS world), or remaining peers need to elect new one (algorithms like Paxos and Raft are designed to target this issue). Anyway, almost whole incoming system traffic goes through the leader. This leads to the "hot spots" in backend: the situation when CPU and IO costs are unevenly distributed across the system. By the way, Raft-based systems have very low write throughput (check etcd and consul limitations if you are interested).
Avoid global state at all. Weaken the guarantees to eventual consistency. Disable the update of files. If someone wants to edit the file, you need to save it as new file. Use the system which is organized as a peer-to-peer network. There is no peer in the cluster that keeps the full track of the system, so there is no single point of failure. This results in high write throughput and nice horizontal scalability.
So now let's discuss the options you've found:
Storing content as BLOBs in databases.
I don't think it's a good option to store files in traditional RDBMS because they provide optimizations for structured data and strong consistency, and you don't need neither of this. Also you'll have difficulties with backups and scaling. People usually don't use RDBMS in this way.
Using GridFS to chunk and store content.
I'm not sure, but it looks like GridFS is built on the top of MongoDB. Again, this is document-oriented database designed to store JSONs, not BLOBs. Also MongoDB had problems with a cluster for many years. MongoDB passed Jepsen tests only in 2017. This may mean that MongoDB cluster is not mature yet. Make performance and stress tests, if you go this way.
Storing content in a file server in directories using a hash and storing the metadata in a database.
This option means that you need to develop object storage on your own. Consider all the problems I've mentioned above.
Using a distributed file system such as GlusterFS or HDFS and storing the file metadata in a database.
I used neither of these solutions, but HDFS looks like overkill, because you get dependent on Hadoop stack. Have no idea about GlusterFS performance. Always consider the design of distributed file systems. If they have some kind of dedicated "metadata" serves, treat it as a single point of failure.
Finally, my thoughts on the solutions that may fit your needs:
Elliptics. This object storage is not well-known outside of the russian part of the Internet, but it's mature and stable, and performance is perfect. It was developed at Yandex (russian search engine) and a lot of Yandex services (like Disk, Mail, Music, Picture hosting and so on) are built on the top of it. I used it in previous project, this may take some time for your ops to get into it, but it's worth it, if you're OK with GPL license.
Ceph. This is real object storage. It's also open source, but it seems that only Red Hat people know how to deploy and maintain it. So get ready to a vendor lock. Also I heard that it have too complicated settings. Never used in production, so don't know about performance.
Minio. This is S3-compatible object storage, under active development at the moment. Never used it in production, but it seems to be well-designed.
You may also check wiki page with the full list of available solutions.
And the last point: I strongly recommend not to use OpenStack Swift (there are lot of reasons why, but first of all, Python is just not good for these purposes).
One probably-relevant question, whose answer I do not readily see in your post, is this:
How often do users actually "modify" the content?
and:
When and if they do, how painful is it if a particular user is served "stale" content?
Personally (and, "categorically speaking"), I prefer to tackle such problems in two stages: (1) identifying the objects to be stored – e.g. using a database as an index; and (2) actually storing them, this being a task that I wish to delegate to "a true file-system, which after all specializes in such things."
A database (it "offhand" seems to me ...) would be a very good way to handle the logical ("as seen by the user") taxonomy of the things which you wish to store, while a distributed filesystem could handle the physical realities of storing the data and actually getting it to where it needs to go, and your application would be in the perfect position to gloss-over all of those messy filesystem details . . .
My understanding was that RavenDb was designed so that if one shard goes down, the other shards can operate without problems.
But recently I was implementing ShardingResolutionStrategy and found out the MetadataShardIdFor method. It is the method where for each document type we can specify what shard to use for storage.
So if I get it right, if the shard where Hilo for specific document type is stored is down, we can not create new documents of this type at other shards (at least autogenerated ids will not work). Or may be I am wrong and Hilo is replicated between shards in some magical way?
Sharding is designed to be independent, but in order to create consistent ids, we need to be able to create them from a consistent store.
Because of that, we separate the notion of splitting data to multiple nodes and HA.
The typical scenario is that the metadata shard is independent, and is running with replicated database that is shared on all sharded nodes. In this fashion, if you lose the metadata shard, you just switch over.
This take advantage on the fact that RavenDB sharding & replication are orthogonal
So I've been trying hard to figure out if NoSQL is really bringing that much value outside of auto-sharding and handling UNSTRUCTURED data.
Assuming I can fit my STRUCTURED data on a single machine OR have an effective 'auto-sharding' feature for SQL, what advantages do any NoSQL options offer? I've determined the following:
Document-based (MongoDB, Couchbase, etc) - Outside of it's 'auto-sharding' capabilities, I'm having a hard time understanding where the benefit is. Linked objects are quite similar to SQL joins, while Embedded objects significantly bloat doc size and causes a challenge regarding to replication (a comment could belong to both a post AND a user, and therefore the data would be redundant). Also, loss of ACID and transactions are a big disadvantage.
Key-value based (Redis, Memcached, etc) - Serves a different use case, ideal for caching but not complex queries
Columnar (Cassandra, HBase, etc ) - Seems that the big advantage here is more how the data is stored on disk, and mostly useful for aggregations rather than general use
Graph (Neo4j, OrientDB, etc) - The most intriguing, the use of both edges and nodes makes for an interesting value-proposition, but mostly useful for highly complex relational data rather than general use.
I can see the advantages of Key-value, Columnar and Graph DBs for specific use cases (Caching, social network relationship mapping, aggregations), but can't see any reason to use something like MongoDB for STRUCTURED data outside of it's 'auto-sharding' capabilities.
If SQL has a similar 'auto-sharding' ability, would SQL be a no-brainer for structured data? Seems to me it would be, but I would like the communities opinion...
NOTE: This is in regards to a typical CRUD application like a Social Network, E-Commerce site, CMS etc.
If you're starting off on a single server, then many advantages of NoSQL go out the window. The biggest advantages to the most popular NoSQL are high availability with less down time. Eventual consistency requirements can lead to performance improvements as well. It really depends on your needs.
Document-based - If your data fits well into a handful of small buckets of data, then a document oriented database. For example, on a classifieds site we have Users, Accounts and Listings as the core data. The bulk of search and display operations are against the Listings alone. With the legacy database we have to do nearly 40 join operations to get the data for a single listing. With NoSQL it's a single query. With NoSQL we can also create indexes against nested data, again with results queried without Joins. In this case, we're actually mirroring data from SQL to MongoDB for purposes of search and display (there are other reasons), with a longer-term migration strategy being worked on now. ElasticSearch, RethinkDB and others are great databases as well. RethinkDB actually takes a very conservative approach to the data, and ElasticSearch's out of the box indexing is second to none.
Key-value store - Caching is an excellent use case here, when you are running a medium to high volume website where data is mostly read, a good caching strategy alone can get you 4-5 times the users handled by a single server. Key-value stores (RocksDB, LevelDB, Redis, etc) are also very good options for Graph data, as individual mapping can be held with subject-predicate-target values which can be very fast for graphing options over the top.
Columnar - Cassandra in particular can be used to distribute significant amounts of load for even single-value lookups. Cassandra's scaling is very linear to the number of servers in use. Great for heavy read and write scenarios. I find this less valuable for live searches, but very good when you have a VERY high load and need to distribute. It takes a lot more planning, and may well not fit your needs. You can tweak settings to suite your CAP needs, and even handle distribution to multiple data centers in the box. NOTE: Most applications do emphatically NOT need this level of use. ElasticSearch may be a better fit in most scenarios you would consider HBase/Hadoop or Cassandra for.
Graph - I'm not as familiar with graph databases, so can't comment here (beyond using a key-value store as underlying option).
Given that you then comment on MongoDB specifically vs SQL ... even if both auto-shard. PostgreSQL in particular has made a lot of strides in terms of getting unstrictured data usable (JSON/JSONB types) not to mention the power you can get from something like PLV8, it's probably the most suited to handling the types of loads you might throw at a document store with the advantages of NoSQL. Where it happens to fall down is that replication, sharding and failover are bolted on solutions not really in the box.
For small to medium loads sharding really isn't the best approach. Most scenarios are mostly read so having a replica-set where you have additional read nodes is usually better when you have 3-5 servers. MongoDB is great in this scenario, the master node is automagically elected, and failover is pretty fast. The only weirdness I've seen is when Azure went down in late 2014, and only one of the servers came up first, the other two were almost 40 minutes later. With replication any given read request can be handled in whole by a single server. Your data structures become simpler, and your chances of data loss are reduced.
Again in my own example above, for a mediums sized classifieds site, the vast majority of data belongs to a single collection... it is searched against, and displayed from that collection. With this use case a document store works much better than structured/normalized data. The way the objects are stored are much closer to their representation in the application. There's less of a cognitive disconnect and it simply works.
The fact is that SQL JOIN operations kill performance, especially when aggregating data across those joins. For a single query for a single user it's fine, even with a dozen of them. When you get to dozens of joins with thousands of simultaneous users, it starts to fall apart. At this point you have several choices...
Caching - caching is always a great approach, and the less often your data changes, the better the approach. This can be anything from a set of memcache/redis instances to using something like MongoDB, RethinkDB or ElasticSearch to hold composite records. The challenge here comes down to updating or invalidating your cached data.
Migrating - migrating your data to a data store that better represents your needs can be a good idea as well. If you need to handle massive writes, or very massive read scenarios no SQL database can keep up. You could NEVER handle the likes of Facebook or Twitter on SQL.
Something in between - As you need to scale it depends on what you are doing and where your pain points are as to what will be the best solution for a given situation. Many developers and administrators fear having data broken up into multiple places, but this is often the best answer. Does your analytical data really need to be in the same place as your core operational data? For that matter do your logins need to be tightly coupled? Are you doing a lot of correlated queries? It really depends.
Personal Opinions Ahead
For me, I like the safety net that SQL provides. Having it as the central store for core data it's my first choice. I tend to treat RDBMS's as dumb storage, I don't like being tied to a given platform. I feel that many people try to over-normalize their data. Often I will add an XML or JSON field to a table so additional pieces of data can be stored without bloating the scheme, specifically if it's unlikely to ever be queried... I'll then have properties in my objects in the application code that store in those fields. A good example may be a payment... if you are currently using one system, or multiple systems (one for CC along with Paypal, Google, Amazon etc) then the details of the transaction really don't affect your records, why create 5+ tables to store this detailed data. You can even use JSON for primary storage and have computed columns derived and persisted from that JSON for broader query capability and indexing where needed. Databases like postgresql and mysql (iirc) offer direct indexing against JSON data as well.
When data is a natural fit for a document store, I say go for it... if the vast majority of your queries are for something that fits better to a single record or collection, denormalize away. Having this as a mirror to your primary data is great.
For write-heavy data you want multiple systems in play... It depends heavily on your needs here... Do you need fast hot-query performance? Go with ElasticSearch. Do you need absolute massive horizontal scale, HBase or Cassandra.
The key take away here is not to be afraid to mix it up... there really isn't a one size fits all. As an aside, I feel that if PostgreSQL comes up with a good in the box (for the open-source version) solution for even just replication and automated fail-over they're in a much better position than most at that point.
I didn't really get into, but feel I should mention that there are a number of SaaS solutions and other providers that offer hybrid SQL systems. You can develop against MySQL/MariaDB locally and deploy to a system with SQL on top of a distributed storage cluster. I still feel that HBase or ElasticSearch are better for logging and analitical data, but the SQL on top solutions are also compelling.
More: http://www.mongodb.com/nosql-explained
Schema-less storage (or schema-free). Ability to modify the storage (basically add new fields to records) without having to modify the storage 'declared' schema. RDBMSs require the explicit declaration of said 'fields' and require explicit modifications to the schema before a new 'field' is saved. A schema-free storage engine allows for fast application changes, just modify the app code to save the extra fields, or rename the fields, or drop fields and be done.
Traditional RDBMS folk consider the schema-free a disadvantage because they argue that on the long run one needs to query the storage and handling the heterogeneous records (some have some fields, some have other fields) makes it difficult to handle. But for a start-up the schema-free is overwhelmingly alluring, as fast iteration and time-to-market is all that matter (and often rightly so).
You asked us to assume that either the data can fit on a single machine, OR your database has an effective auto-sharding feature.
Going with the assumption that your SQL data has an auto-sharding feature, that means you're talking about running a cluster. Any time you're running a cluster of machines you have to worry about fault-tolerance.
For example, let's say you're using the simplest approach of sharding your data by application function, and are storing all of your user account data on server A and your product catalog on server B.
Is it acceptable to your business if server A goes down and none of your users can login?
Is it acceptable to your business if server B goes down and no one can buy things?
If not, you need to worry about setting up data replication and high-availability failover. Doable, but not pleasant or easy for SQL databases. Other types of sharding strategies (key, lookup service, etc) have the same challenges.
Many NoSQL databases will automatically handle replication and failovers. Some will do it out of the box, with very little configuration. That's a huge benefit from an operational point of view.
Full disclosure: I'm an engineer at FoundationDB, a NoSQL database that automatically handles sharding, replication, and fail-over with very little configuration. It also has a SQL layer so you you don't have to give up structured data.
I want to use Redis as a database, not a cache. From my (limited) understanding, Redis is an in-memory datastore. What are the risks of using Redis, and how can I mitigate them?
You can use Redis as an authoritative store in a number of different ways:
Turn on AOF (Append-only File store) see AOF docs. This will keep a log of all Redis commands made against your dataset in real-time.
Run Redis using Master-Slave replication see replication docs. This will allow you to provide high-availability if one of your instances fails.
If you're running on something like EC2 you can EBS back your Redis partition to provide another layer of protection against instance failure.
On the horizon is Redis Cluster - this is specifically designed as a way to run Redis in a way that should help with HA and scalability. However, this won't appear for at least another six months or so.
Redis is an in-memory store which can also write the data back to disc. You can specify how many times to do a fsync to make redis safer(but also slower => trade-off) .
But still I am not certain if redis is in state yet to really store (mission) critical data in it (yet?). If for example it is not a huge problem when 1 more tweets(twitter.com) or something similiar get losts then I would certainly use redis. There is also a lot of information available about persistence at redis's own website.
You should also be aware of some persistence problems which could occur by reading antirez(redis maintainers) blog article. You should read his blog because he has some interesting articles.
I would like to share a few things that we have learned by using Redis as a primary Database in our service. We choose Redis since we had data that could not be partitioned. We wanted to get the best performance we could get out of one box
Pros:
Redis was unbeatable in raw performance. We got 10K transactions per second out of the box (Note that one transaction involved multiple Redis commands). We were able to hit a rate of 25K+ transactions per second after a few optimizations, along with LUA scripts. So when it comes to performance per box, Redis is unmatched.
Redis is very simple to setup and has a very small learning curve as opposed to other SQL and NoSQL datastores.
Cons:
Redis supports only few primitive Data Structures like Hashes, Sets, Lists etc. and operations on these Data Structures. These are more than sufficient when you are using Redis as a cache, but if you want to use Redis as a full fledged primary data store, you will feel constrained. We had a tough time modelling our data requirements using these simple types.
The biggest problem we have seen with Redis was the lack of flexibility. Once you have solutioned the structure of your data, any modifications to storage requirements or access patterns virtually requires re-thinking of the entire solution. Not sure if this is the case with all NoSQL data stores though (I have heard MongoDB is more flexible, but haven't used it myself)
Since Redis is single threaded, CPU utilization is very low. You can't put multiple Redis instances on the same machine to improve CPU utilization as they will compete for the same disk, making disk as the bottleneck.
Lack of horizontal scalability is a problem as mentioned by other answers.
As Redis is an in-memory storage, you cannot store large data that won't fit you machine's memory size. Redis usually work very bad when the data it stores is larger than 1/3 of the RAM size. So, this is the fatal limitation of using Redis as a database.
Certainly, you can distribute you big data into several Redis instances, but you have to do it all on your own manually. The operation usually be done like this(assuming you have only 1 instance from start):
Use its master-slave mechanism to replicate data to the second machine, Now you have 2 copies of the same data.
Cut off the connection between master and slave.
Delete the first half(split by hashing, etc) of data on the first machine, and delete the second half of data on the second machine.
Tell all clients(PHP, C, etc...) to operate on the first machine if the specified keys are on that machine, otherwise operate on the second machine.
This is the way how Redis scales! You also have to stop your service to prevent any writes during the migration.
To the expierence we encounter, we have this conclusion to Redis: Redis is not the right choice to store more than 30G data, Redis is not scalable, Redis is quite suitable for prototype development.
We later find an alternative to Redis, that is SSDB(https://github.com/ideawu/ssdb), a leveldb server that supports nearly all the APIs of Redis, it is suitable for storing more than 1TB of data, that only depends on the size of you harddisk.
Redis is a database, that means we can use it for persisting information for any kind of app, information like user accounts, blog posts, comments and so on. After storing information we can retrieve it later on by writing queries.
Now this behavior is similar to just about every other database, but what is the difference? Or rather why would we use it over any other database?
Redis is fast.
Redis is not fast because it's written in a special programming language or anything like that, it's fast because all data is stored in-memory.
Most databases store all their information between both the memory of a computer and the hard drive. Accessing data in-memory is fast, but getting it stored on a hard disk is relatively slow.
So rather than storing memory in hard disk, Redis decided to store it in memory.
Now, the downside to this is that working with data that is larger than the amount of memory your computer has, that is not going to work.
That may sound like a tremendous problem, but Redis has clear strategies for working around this limitation.
The above is just the first reason why Redis is so fast.
The second reason is that Redis stores all of its data or rather organizes all of its data in simple data structures such as Doubly Linked Lists, Sorted Sets and so on.
These data structures have well-known and well-understood performance characteristics. So as developers we can decide exactly how our information is organized and how to efficiently query data.
It's also very fast because Redis is simple in nature, it's not feature heavy; feature heavy datastores like Postgres have performance penalties.
So to use Redis as a database you have to know how to store in limited space, you have to know how to organize it into these simple data structures mentioned above and you have to understand how to work around the limited feature set.
So as far as mitigating risks, the way you start to do that is to start to think Redis Design Methodology and not SQL Database Design Methodology. What do I mean?
So instead of, step 1. Put the data in tables, step 2. figure out how we will query it.
With Redis it's more:
Step 1. Figure out what queries we need to answer.
Step 2. Structure data to best answer those queries.
I am considering the option of neo4j for some of the new projects I am working for. For the given data needs (inherently graph based) neo4j fits well and a quick prototype is giving good response time for me. What I want to understand is how to scale a neo4j deployment. Specifically:
How do I shard my data across neo4j deployments. Since neo4j is deployed on a single machine, there is a limit to how much data I can store in a single machine and hence I would like to know how to distribute it. Clearly if I split it on users, then relationships between disconnected users (across shards) cannot be maintained.
How do I replicate the neo4j data? I am potentially thinking of putting up a sql-like-setup with masters used for write and slaves used for reads so that we can both scale up our potentially readers and writers, but also have a backup of our data in real time. I understand that all the neo4j data is stored in a filesystem - which is inherently non-replicatable. Is there a way I can do it here? Perhaps, something akin to a mysql bin log?
sharding is as of now not handled by Neo4j itself, but by the domain, much as you describe. Neo4j 2.0 is going to target that problem.
For replication, Online Backup is working and real High Availability with Master failover is in the works, using ZooKeeper to track the cluster nodes and elect new masters, etc.
Any more details on your app sharding requirements? What domain etc?