Commonly used technology¶
High availability can be achieved only on system level, while both hardware and software components can contribute to the system level availability. This document lists the most common hardware and software technologies that can be used to build a highly available system.
Hardware¶
Using different technologies to enable high availability on the hardware level provides a good basis to build a high available system. The next chapters discuss the most common technologies used in this field.
Redundant switches¶
Network switches are single point of failures as networking is critical to operate all other basic domains of the infrastructure, like compute and storage. Network switches need to be able to forward the network traffic and be able to forward the traffic to a working next hop. For these reasons consider the following two factors when making a network switch redundant:
The network switch itself should synchronize its internal state to a redundant switch either in active/active or active/passive way.
The network topology should be designed in a way that the network router can use at least two paths in every critical direction.
Bonded interfaces¶
Bonded interfaces are two independent physical network interfaces handled as one interface in active/passive or in active/active redundancy mode. In active/passive mode, if an error happens in the active network interface or in the remote end of the interface, the interfaces are switched over. In active/active mode, when an error happens in an interface or in the remote end of an interface, then the interface is marked as unavailable and ceases to be used.
Load balancers¶
Physical load balancers are special routers which direct the traffic in different directions based on a set of rules. Load balancers can be in redundant mode similarly to the physical switches. Load balancers are also important for distributing the traffic to the different active/active components of the system.
Storage¶
Physical storage high availability can be achieved with different scopes:
High availability within a hardware unit with redundant disks (mostly organized into different RAID configurations), redundant control components, redundant I/O interfaces and redundant power supply.
System level high availability with redundant hardware units with data replication.
Software¶
HAproxy¶
HAProxy provides a fast and reliable HTTP reverse proxy and load balancer for TCP or HTTP applications. It is particularly suited for web crawling under very high loads while needing persistence or Layer 7 processing. It realistically supports tens of thousands of connections with recent hardware.
Note
Ensure your HAProxy installation is not a single point of failure, it is advisable to have multiple HAProxy instances running.
You can also ensure the availability by other means, using Keepalived or Pacemaker.
Alternatively, you can use a commercial load balancer, which is hardware or software. We recommend a hardware load balancer as it generally has good performance.
For detailed instructions about installing HAProxy on your nodes, see the HAProxy official documentation.
keepalived¶
keepalived is a routing software that provides facilities for load balancing and high-availability to Linux system and Linux based infrastructures.
Keepalived implements a set of checkers to dynamically and adaptively maintain and manage loadbalanced server pool according their health.
The keepalived daemon can be used to monitor services or systems and to automatically failover to a standby if problems occur.
Pacemaker¶
Pacemaker cluster stack is a state-of-the-art high availability and load balancing stack for the Linux platform. Pacemaker is used to make OpenStack infrastructure highly available.
Pacemaker relies on the Corosync messaging layer for reliable cluster communications. Corosync implements the Totem single-ring ordering and membership protocol. It also provides UDP and InfiniBand based messaging, quorum, and cluster membership to Pacemaker.
Pacemaker does not inherently understand the applications it manages. Instead, it relies on resource agents (RAs) that are scripts that encapsulate the knowledge of how to start, stop, and check the health of each application managed by the cluster.
These agents must conform to one of the OCF, SysV Init, Upstart, or Systemd standards.
Pacemaker ships with a large set of OCF agents (such as those managing MySQL databases, virtual IP addresses, and RabbitMQ), but can also use any agents already installed on your system and can be extended with your own (see the developer guide).
