Trying to figure out what kind of machine to requisition for a MongoDB Arbiter so I can set up a replica set. All it needs to do is vote when a machine goes down, right? Could there possibly be any problem with using a micro EC2 instance for this?
I would not use Amazon Micro for an Arbiter as Micro instances can have long lags in processing which could trigger the replication threshold for non-response. In this case you will go into Secondary only ready only mode and your app can't do any writes as once the arbiter has voted it can't revote.
Go with at least m1.small as you are guaranteed computing power or as someone else suggested throw it as a share on another instance.
Please be advised that sharing with other instances complicates your error troubleshooting so it's much easier to just put it on it's own instance.
FYI - We run mongo replicaset for mission critical app on EC2 with arbiter and have also worked with Mongo consultants to come to this conclusion.
You are correct - Arbiters are nodes in a replica set that only participate in elections. Further information can be found here:
http://www.mongodb.org/display/DOCS/Adding+an+Arbiter
A micro EC2 instance should be fine for this if you do need one. .
Related
I'm fairly new to mongodb, just a couple of months. I just converted my mongodb database to support a secondary replica set so I can watch collections. I only added one secondary which I'm guessing now may not be the best after reading you should create an odd number, but it is a localhost on one machine. I went through the instructions, got replication working fine for for half a day running my programs. But for some reason recently it has switched the database for port 27017 from primary to secondary. Primary was previously on localhost:27017 and secondary was on localhost:27027. Now my normal program can't connect to localhost:27017 without an error, which I believe it is because it is a secondary replica set now when it was primary before, assuming you can only connect to a primary. Here is the error msg.
Exception in thread "main" com.mongodb.MongoNotPrimaryException: Command failed with error 10107 (NotWritablePrimary): 'not master' on server localhost:27017.
I'm perplexed why mongodb switched the replica set primary in the first place. I doubt an error occurred, but certainly possible, but I haven't had a single "localhost" error in months of development.
For now, ideally how would I switch 27017 back to be the primary. How do I do that so my existing programs can function again?
Eventually when in production, what is the best methodology to handle this, assuming a lookup to a DNS entry to an ip address and suddenly the primary gets changed because of a fail over?
Given question 3 is a bit more involved, is there something I can do in my development environment to better simulate a production environment.
I use StackOverflow extensively but this is my first post so thanks for anyone who can provide advice.
Without knowing more about the replica configuration and circumstances of the switch over I'm not sure anyone could confidently answer question 1 but it may not be important compared to question 3.
When you want to manually switch the primary you can manipulate the priority settings:
https://docs.mongodb.com/manual/tutorial/force-member-to-be-primary/
Or run manual commands to freeze or step down the current primary:
https://docs.mongodb.com/manual/tutorial/force-member-to-be-primary/#force-a-member-to-be-primary-using-database-commands
The safest option is to ensure your application is aware of all replicas in the replica set. Then when you have these situations where something unexpected has happened the application will fail over to a writable db without any issues.
https://mongodb.github.io/mongo-java-driver/3.4/driver/tutorials/connect-to-mongodb/#connect-to-a-replica-set
I can only suggest setting up some VMs or containers as replica set members to better represent a production environment.
https://hub.docker.com/_/mongo
I was able to solve the problem by using the connect string which comprised both replica sets, which I was unaware I needed to do. Such as for java:
mongoClient = MongoClients.create("mongodb://localhost:27017,localhost:27027");
This also worked for Mongo Compass so I was able to connect to the secondary database. I didn't know you needed to provide paths to all replica sets when trying to connect, but in retrospect makes goods sense if something is down.
If you need a replica set for testing, you can create a single-node RS. Follow the instructions for creating a RS but only add one node.
We are currently setting up our mongodb environment for production. At the moment we only have one dedicated mongodb database server. We will expand this in the near future with a 2nd server and I already indicated to the management that for the ideal situation we should get a 3rd server as well.
Since I already know we're going to use sharding and replication in the near future I want to be prepared for it.
The idea I have now is to start now with the Development Configuration (as mongo's documentation names it).
Whenever our second server comes available I would like to expand this setup to a configuration with 2 configuration servers en 2 shards (replica sets).
And of course when our third server comes available have the fully functional sharded cluster configuration.
While reading mongo's documentation I was getting triggered by the note that de Development setup should not be used in production.
MongoDb Development Configuration
Keeping in mind that we will add more servers soon, would it be a bad idea to already configure the Development Configuration already so we can easily add the 2nd server to the cluster when it comes available?
After setting up the 'development sharded setup' I've found my anwser. Of course i'm happy to share in case anybody runs into the same questions as I do when starting with this.
In my case, it was ok to start with the development setup untill my new servers arrived. It was a temporary situation and when my new servers arived I was able to easily expand my replicasets. There are a number of reasons why this isn't adviced for production:
To state the obvious, there is no replication yet. Since I was running shards on one machine there is a single point of failure. If the machine, or one node goes down, the cluster won't work anymore.
Now this part is interesting. After I added a second server, I did have primary and secondary nodes. Primary nodes were used for writing and secondary for reading. I've eliminated the issue that there was no replication AND my data had a higher availability. However, I noticed with the 2-member replica sets, if one member of the replicaset went down (even is this was a secondary), the primary stepped down to a secondary node as well. This had to do with the voting mechanism that MongoDb uses. See Markus' more detailed answer on this.. Since there are no more primaries in the replicaset, my cluster won't function anymore. Now, if i were to use an arbiter I could eliminate this problem as well.
When you have a 3-member replicataset, automatic failover kicks in. Whenever a node goes down, another primary is assigned automatically and the cluster will continue performing as before.
During my tests I also got to a point where one of my MongoD.exe instances stopped working due to a "Out of memory exception". I was running a cluster with 3 replicasets, meaning every machine had at least 4 mongod.exe processes running (3 for the replicaset shards and one for the configuration server replicaset). Besides having a query which wasn't optimized yet I also noticed that the WiredTiger storage engine by default can use up to 50% of ram minus one gigabyte. Perhaps it wasn't the best choise to have multiple replicaset-shards on one machine but I was able to eliminate the problem by capping the wiredtiger memory usage.
I hope this answer helps anybody who's starting to set up replication and sharding for MongoDb.
I would like to have Mongo on a server, then to replicate this onto a laptop.
The laptop is needed to leave the network and still be able to read/write, and once back on the network sync these changes with the primary.
At the same time I need the VM (Primary), to still be accessible (read/write).
So when each device is not talking to each other then for them to make themselves primary.
I have set up a very basic replica, Primary on a VM and the Secondary on the machine running the VM. In all examples I have seen it recommends having 3 servers for the replica, but I only need 2!
A couple of questions:
Is this possible with Mongo? If not then any suggestions!
When I turn off the Network adapter on the VM(Primary), the secondary doesn't seem to want to become the primary.
Is it possible to run 2 instance of Mongo, and then use the other instance as the 3 member of the replica.
Any advice would be great, thanks.
MongoDB always needs even number servers. in your case it should be like one primary one secondary and one arbiter. arbiter server instance is responsible for electing new server as primary when current primary goes down.
Here is a newbie trying to play around Mongodb. I am trying to demonstrate scaling in my class, meaning, I need to show that I have 2 instances of mongoDB up and running and I need to replicate them, set one as master and the other as secondary.
Can any of you suggest me a simple way to demonstrate that if primary/master fails the slave/secondary comes up as the master?
Please keep it as simple as possible as I am teaching to a fairly beginners of MongoDB
MongoDB replica sets are not master/slave. In order to achieve automatic failover you need to have a majority of nodes in the replica set able to elect a new primary. The minimum number of nodes in your replica set should be 3, which can either be 3 data-bearing nodes or 2 data-bearing nodes and an arbiter, which is a node that votes in elections.
A demo using replication alone is more about failover and redundancy than scaling (better demo'd with sharding).
If you want a very simple (and non-production) way to stand up a replica set or sharded cluster in a development environment, I would suggest using the mlaunch script which is part of mtools.
For example, to create a 3-node replica set with an arbiter:
mlaunch --replicaset --nodes 2 --arbiter
To create a sharded cluster with 3 shards backed by a replica set (plus mongos and config server):
mlaunch --replicaset --sharded 3
As mentioned in the other comments here, the free MMS Monitoring service is a good way to visualise activity in your MongoDB deployment, and you can use db.shutdownServer() to shutdown specific nodes to see the outcome.
The easiest way would be to set up the MongoDB monitoring service. Stop the MongoD process on one and watch the other take over. But, use replica sets rather than master/secondary as they are the recommended approach.
Actually, it is pretty easy
Set up a replica set with 2 "normal" mongods and an arbiter
Connect to both of the normal mongods using mongo
Show the output of rs.status(). (Note the selffield)
Shut down the current primary
Show the output of rs.status() again and again, until the former secondary is elected primary
Another option would be to write a simple java app which utilizes the driver, put it in an infinite loop which writes one entry every second and puts out the number of objects in the database. Catch exceptions and write out that a problem occurred. Start the sharded cluster, then start your application. Shut down the primary while the program is running. during the elections, there may be exceptions be thrown. As soon as the former secondary is elected primary, the document count should start to rise again.
I need the proper way of failover mechanism for mongodb on aws ec2. I know failover can be accomplished by replica sets, but what is the best way to fire a new mongo installed ubuntu-ec2 ami node and add it to replica set again automatically (with zero manual operation) and return the replica set to it's proper state ?
EBS has some problems, but if I use local instance storage, I will lost the dead nodes data, but does the replica got all the master data and so is replaca is enough to recover everthing (on mongo 1.8 with journaling), or do I have to use only EBS ?
How should I start mongo instances, If I should start with repair option, how can I sperate node's first run from failover restart ?
Regards,
The easiest way to bring up new nodes is to bring up a new node with a recent backup.
So now it's a question of how you do your backup and how you restore from the backup quickly.
The MongoDB site has a write up for backups (in general) and backups on EC2 specifically. There's also a write-up for adding a new set member.
You can do this with instance storage or EBS drives, but you'll need different strategies for each. There's really no single way to do this, so I would check out the docs I've linked to for a primer.
Highly recommend reading Sean Coates' article on mutli-node MongoDB Elections, failover and AWS - specifically, the subtlety on distributed arbiter nodes (e.g., make sure to give yourself a voting majority when an AZ goes down). A similar recommendation can be found in a comment on this (now-closed) MongoDB vs. Cassandra thread.