Lite HA 3.0
Now, more and more embedded system also has high availability requirements. For embedded market, Skyge Soft release embedded high availability solution LiteHA 3.0. This collection of product is a fast and light weight high availability solution based on Linux platform.
Brief Introduction Of SkygeHA LiteHA 3.0 Features
Skyge LiteHA 3.0 focus on protecting critical business in embedded system, providing 1-1 hot standby system. When master node fails, standby node take over the master’s role and resources to continue providing services. That will help customer to provide 24 * 7 non-stop services based on Skyge’s solution.
Skyge LiteHA 3.0 provides powerful agent mechanism to adapt to different services, LiteHA has provided build-in standard agents for popular services, such as MySQL,Apache and so on.
For customer’s typical services, Skyge Soft provide technical manuals to help customer to write their own agents. Skyge Soft also provide services to help building agent freely.
Skyge HA integrated with DRBD seamlessly, DRBD provide a network disk mirror function. With DRBD, Skyge HA doesn’t need a shared disk array in 1-1 high availability environments, DRBD can form disks on two nodes as one disk, when any data written on master node, the data is also written on standby node at the same time. When failover switch happen, applications can get data from standby node and continue to provide service.
Multi-channels To Avoid Split-Brain
Skyge LiteHA 3.0 can set up multiple channels between two nodes to avoid split-brain problem, the more channels configured, more safety system can gain. For some extreme situation that all the channels break, Skyge HA provides third party IP mechanism, this will reboot node to avoid split-brain when node finds itself is in isolated state. Skyge HA can configure more than one third party IP to enhance this feature.
Preferential Node And Intelligent Take-back Policy
Skyge LiteHA support preferential node mechanism. User can identify a preferential node for some services, when preferential node fails, services are switched to standby node. Services will be pulled back to preferential node when it’s fixed and online again. Preferential node feature can be used in such scenario, the master node are more powerful than standby node, user can use preferential node feature to allow more services running on master node than standby node.
Overall Failure Detection
Based on years accumulated cases study, Skyge LiteHA 3.0 provide overall failure detection, there are over 200 build-in failure models which provide comprehensive analysis for system. Those failure detection includes disk failure detection, network failure detection, memory failure detection, hardware failure and so on.
Skyge LiteHA 3.0 provide comprehensive log function, both master and standby nodes will record useful info in log file. Log record includes very detailed system info, running state, actions, timestamp and so on. those information is very useful for post-mortem analysis when failure happen and failover switch actions taken.
Skyge Skyge HA log system has different levels, user can change log level according to requirements. By default, Skyge HA log level is high level, which will output almost all the details.
Log system is divided into several catalogs, they are service log, file check log, script execution log, those catalogs can help users to identify problem more clearly.