Changes to the same data concurrently in multiple datacenters has to be resolved – conflict resolution – to be discussed later.So it’s clear right? Multi-leader all the things? Hint: No! Problems with Multi-Leader Replication Usually a multi-leader replication is more capable of handling network issues as there are multiple data centers handling the writes – therefore a major issue in one datacenter doesn’t cause everything to take a dive.In multi-leader, the other datacenters can continue taking writes and will catch back up when a new leader is selected in the datacenter where the failure occurred.Fault tolerance – in single-leader, everything is on pause while a new leader is elected.

The synchronization of that data across datacenters can happen asynchronously making the system feel faster overall.Performance – because writes can occur in each datacenter without having to go through a single datacenter, latency can be greatly reduced in multi-leader Sort of a chained single-leader replication setupĬomparing Single-Leader vs Multi-Leader Replication.However, each leader would be followers to the other datacenter leaders An interesting approach is for each datacenter to have a leader and followers…similar to the single leader.When you have multiple leaders you would typically have a leader in each datacenter.you don’t want to have multiple leaders within the same datacenter because the complexity outweighs the benefits. Many database systems that support single-leader replication can be taken a step further to make them mulit-leader.Each leader acts as followers to other leaders.Most things behave just like single-leader replication.Allows more than one node to receive writes.Should the old leader come back then…we have a problem. There was a single point of failure for writes, and it could take time to figure out the new leader. Single leader replication had some problems. Replication isn’t as easy as it sounds, is it? Multi-Leader Rep…lication We’ll talk more about this problem in a future episode, but the short answer is to make sure that causal data gets sent to the same “partition”.Consistent Prefix Reads: Think about causal data…an order is placed, and then the order is shipped…but what if we had writes going to more than one spot and you query the order is shipped.but nothing was placed? (We didn’t have this problem with a Single Replica).Another possibility – ensure that the reads are always coming from the same replica.One way to do this – keep a timestamp of the most recent read data, discard any reads older than that…you may get errors, but you won’t see data older than you’ve already seen. Monotonic Reads: a guarantee of sorts that ensures you won’t see data moving backwards in time.You make sure to route all devices for a user the same way.Centralize MetaData (1 leader to read from for everything).But…what if the user is using multiple devices?.Client can keep a timestamp of it’s most recent write, then only allow reads from a replica that has that timestamp (logical clocks keep problems with clock synchronization at bay here).Read from a leader/In Sync Replica for some period of time after a change.Read important data from a leader if a change has been made OR if the data is known to only be changeable by that particular user (user profile).In general this ensures that the user who wrote the data will get the same data back – other users may get stale version of the data.Read You Writes Consistency refers to an attempt to read significant data from leader or in sync replicas by the user that submitted the data.Techniques for mitigation replication lag There are a couple techniques that can help with this problem.Some common use cases make this particularly bad, like a user updating some information…they often expect to see that change afterwards.“Eventual” is purposely vague – could be a few seconds, could be an hour.You have to design your app knowing that followers will eventually catch up – “ eventual consistency“.Potentially very stale (it does dial up the availability and perhaps performance) The problem is if you allow async writes, then your data can be stale.The upside is that your data is consistent.The more nodes the more likely there will be a problem with one or more.That doesn’t work well with sync writes.the more followers, the higher the latency.If your application is read heavy, then you can add followers to increase your scalability.When you’re talking about single or multi-leader replication, remember all writes go through leaders.Talking about Multi-Leader Replication Replication Recap and Latency

Iterm2 split window