It may seem strange to some that journaling was added as late as version 1.8 to MongoDB. Journaling is only now the default for 64 bit OS for MongoDB 2.0. Prior to that, you typically used replication to make sure write operations were copied to a replica before proceeding if the data was very important. The thought being that one server might go down, but two servers are very unlikely to go down at the same time. Unless somebody backs a truck over a high voltage utility poll causing all of your air conditioning equipment to stop working long enough for all of your servers to overheat at once, but that never happens (it happened to Rackspace and Amazon). And if you were worried about this, you would have replication across availability zones, but I digress.
At one point MongoDB did not have single server durability, now it does with addition of journaling. But, this is far from a moot point. The general thought from MongoDB community was and maybe still is that to achieve Web Scale, durability was thing of the past. After allmemory is the new disk. If you could get the data on second server or two, then the chances of them all going down at once is very, very low. How often do servers go down these days? What are the chances of two servers going down at once? The general thought from MongoDB community was (is?) durability is overvalued and was just not Web Scale. Whether this is a valid point or not, there was much fun made about this at MongoDB's expense (rated R, Mature 17+).
As you recall MongoDB uses memory mapped file for its storage engine so it could be a while for the data in memory to get synced to disk by the operating system. Thus if you did have several machines go down at once (which should be very rare), complete recoverability would be impossible. There were workaround with tradeoffs, for example to get around this (now non issue) or minimize this issue, you could force MongoDB to do an fsync of the data in memory to the file system, but as you guessed even with a RAID level four and a really awesome server that can get slow quick. The moral of the story is MongoDB has journaling as well as many other options so you can decide what the best engineering tradeoff in data safety, raw speed and scalability. You get to pick. Choose wisely.
The reality is that no solution offers "complete" reliability, and if you are willing to allow for some loss (which you can with some data), you can get enormous improvements in speed and scale. Let's face it your virtual farm game data is just not as important as Wells Fargo's bank transactions. I know your mom will get upset when she loses the virtual tractor she bought for her virtual farm with her virtual money, but unless she pays real money she will likely get over it. I've lost a few posts on twitter over the years, and I have not sued once. If your servers have an uptime of 99 percent and you block/replicate to three servers than the probability of them all going down at once is (0.000001) so the probability of them all going down is 1 in 1,000,000. Of course uptime of modern operating systems (Linux) is much higher than this so one in 100,000,000 or more is possible with just three servers. Amazon EC2 offers discounts if they can't maintain an SLA of 99.95% (other cloud providers have even higher SLAs). If you were worried about geographic problems you could replicate to another availability zone or geographic area connected with a high-speed WAN. How much speed and reliability do you need? How much money do you have?
An article on when to use MongoDB journaling versus older recommendations will be a welcome addition. Generally it seems journaling is mostly a requirement for very sensitive financial data and single server solutions. Your results may vary, and don't trust my math, it has been a few years since I got a B+ in statistics, and I am no expert on SLA of modern commodity servers (the above was just spit balling).
If you have ever used a single non-clustered RDBMS system for a production system that relied on frequent backups and transaction log (journaling) for data safety, raise your hand. Ok, if you raised your hand, then you just may not need autosharding or replica sets. To start with MongoDB, just use a single server with journaling turned on. If you require speed, you can configure MongoDB journaling to batch writes to the journal (which is the default). This is a good model to start out with and probably very much like quite a few application you already worked on (assuming that most application don't need high availability). The difference is, of course, if later your application deemed to need high availability, read scalability, or write scalability, MongoDB has your covered. Also setting up high availability seems easier on MongoDB than other more established solutions.
Figure 3: Simple setup with journaling and single server ok for a lot of applications
If you can afford two other servers and your app reads more than it writes, you can get improved high availability and increased read scalability with replica sets. If your application is write intensive then you might need autosharding. The point is you don't have to be Facebook or Twitter to use MongoDB. You can even be working on a one-off dinky application. MongoDB scales down as well as up.
Replica sets are good for failover and speeding up reads, but to speed up writes, you need autosharding. According to a talk by Roger Bodamer on Scaling with MongoDB, 90% of projects do not need autosharding. Conversely almost all projects will benefit from replication and high availability provided by replica sets. Also once MongoDB improves its concurrency in version 2.2 and beyond, it may be the case that 97% of projects don't need autosharding.
Sharding allows MongoDB to scale horizontally. Sharding is also called partitioning. You partition each of your servers a portion of the data to hold or the system does this for you. MongoDB can automatically change partitions for optimal data distribution and load balancing, and it allows you to elastically add new nodes (MongoDB instances). How to setup autosharding is beyond the scope of this introductory article. Autosharding can support automatic failover (along with replica sets). There is no single point of failure. Remember 90% of deployments don’t need sharding, but if you do need scalable writes (apps like Foursquare, Twitter, etc.), then autosharding was designed to work with minimal impact on your client code.
There are three main process actors for autosharding: mongod (database daemon), mongos, and the client driver library. Each mongod instance gets a shard. Mongod is the process that manages databases, and collections. Mongos is a router, it routes writes to the correct mongod instance for autosharding. Mongos also handles looking for which shards will have data for a query. To the client driver, mongos looks like a mongod process more or less (autosharding is transparent to the client drivers).
Autosharding increases write and read throughput, and helps with scale out. Replica sets are for high availability and read throughput. You can combine them as shown in figure 5.
Figure 5: MongoDB Autosharding plus Replica Sets for scalable reads, scalable writes, and high availability
You shard on an indexed field in a document. Mongos collaborates with config servers(mongod instances acting as config servers), which have the shard topology (where do the key ranges live). Shards are just normal mongod instances. Config servers hold meta-data about the cluster and are also mongodb instances.
Shards are further broken down into 64 MB chunks called chunks. A chunk is 64 MB worth of documents for a collection. Config servers hold which shard the chunks live in. The autosharding happens by moving these chunks around and distributing them into individual shards. The mongos processes have a balancer routine that wakes up so often, it checks to see how many chunks a particular shard has. If a particular shard has too many chunks (nine more chunks than another shard), then mongos starts to move data from one shard to another to balance the data capacity amongst the shards. Once the data is moved then the config servers are updated in a two phase commit (updates to shard topology are only allowed if all three config servers are up).
The config servers contain a versioned shard topology and are the gatekeeper for autosharding balancing. This topology maps which shard has which keys. The config servers are like DNS server for shards. The mongos process uses config servers to find where shard keys live. Mongod instances are shards that can be replicated using replica sets for high availability. Mongos and config server processes do not need to be on their own server and can live on a primary box of a replica set for example. For sharding you need at least three config servers, and shard topologies cannot change unless all three are up at the same time. This ensures consistency of the shard topology. The full autosharding topology is show in figure 6. An excellent talk on the internals of MongoDB sharding was done by Kristina Chodorow, author of Scaling MongoDB, at OSCON 2011 if you would like to know more.
If you would like to learn more about MongoDB consider the following resources:
Figure 6: MongoDB Autosharding full topology for large deployment including Replica Sets, Mongos routers, Mongod Instance, and Config Servers
Mammatus Blog >