Linux High Availability Clustering

Start and stop cluster services on individual nodes or across all nodes, and enable or disable auto-join behavior to manage node readiness in a Pacemaker Corosync cluster.

Learn to add a new node to a running high availability cluster online, without downtime, using pacemaker and fencing agents with firewall configuration and authenticated cluster node addition.

Configure fencing for a newly added node in Linux high availability clustering using pxe stonith and the fence_xvm agent, copying the secret key to the target and verifying the reboot.

Remove a node from a running four-node cluster with pxe cluster node remove and delete its fencing via pxe stonith delete, then use standby commands to suspend resources during maintenance.

Understand quorum operations through majority voting, how many votes keep a cluster online, and how fencing interacts to maintain operational services when nodes fail.

Practice lab demonstrates managing quorum in a four-node Linux high availability cluster using Corosync quorum, showing how votes determine cluster operation as nodes fail and rejoin.

Configure quorum with the wait_for_all option to require all cluster members online before calculating quorum. Demonstrate how this prevents fence race by enabling the switch in a four-node setup.

Illustrate quorum calculation and how auto_tie_breaker keeps a two-node cluster operational with 50% quorum. Demonstrate split brain and how the lowest node ID wins quorum when nodes cannot communicate.

learn to create and configure high-availability resources in a two-node cluster using pacemaker, grouping ip, file system, and web service resources for reliable failover.

Configure a resource group in a three-node cluster to run a clustered Apache web service, using a floating IP, NFS file system, and httpd resource, with ordered start and failover.

Learn to manage cluster resources and resource groups with start, stop, disable, move, and ban operations across nodes A, B, and C using the high availability add-on.

Inspect and configure cluster logging to troubleshoot high availability issues, using corosync and pacemaker logs, adjust log file locations, enable debugging, and restart the cluster to apply changes.

Troubleshoot resource failures by checking logs, fail counts, and configuration files, then use debug mode and update the resource to restore service.

Troubleshoot cluster network issues in a three-node Linux high availability cluster by pinging the offline node, verifying SSH, and inspecting Corosync quorum and logs to restore connectivity.

Configure an active/passive NFS resource group in a three-node pacemaker cluster with shared iSCSI storage, exporting a ZFS file system to clients and ensuring seamless failover.

Configure an active-passive NFS resource group on a three-node cluster to share iSCSI target storage, create a file system, export it via NFS, and enable failover with a floating IP.

Learn to set order constraints to define start sequence, location constraints to assign resources to specific nodes, and co-location constraints to control same-node placement for resource groups.

Learn to configure order and location constraints in a two-node cluster, guiding startup order and node preferences for resources like NFS and web server, using constraint order and constraint location.

Set location constraints and resource stickiness in linux high availability clustering to keep the web server resource group on node A during cluster restarts and avoid outages.

Explore co-location constraints in pacemaker, forcing or preventing two resources or resource groups from running on the same node, using A with B and minus infinity to never co-locate.

Explore two-node cluster challenges and how quorum prevents node failure, using the two_node_node setting to reduce required votes so a failed node doesn't crash the cluster.

Split brain in a two-node Linux cluster happens when nodes lose communication, each thinks it is online and activates shared resources, risking data corruption; fencing prevents by rebooting peers.

Learn how fencing and fence racing prevent split brain in two-node Linux high availability clusters. Apply delayed fencing to break the race and keep the web service on one node.

Explains that in a two-node cluster, the system waits for both nodes to start before forming quorum, enabling single-node operation if one node is down, via two_node and wait_for_all settings.

Configure and access storage on a Linux server using iSCSI targets and initiators over IP networks, sharing storage via LUNs and SCSI commands.

Learn the core iSCSI fundamentals, including initiators and targets, iqn naming, and access control with ACLs, plus discovery and login steps, all within the ESXi context.

Explore iSCSI component terminology, including iSCSI qualified names, initiators and targets, LUNs, login and discovery, and portal concepts, and learn how to configure ESXi iSCSI targets and initiators.

Log into the target server, install the target CLI package, and manage target services with the Stargate Seela utility to configure megastores, block, file, and ramdisk storage objects.

Create and expose a block store with iSCSI on the target, define a lun and acl, and configure the client initiator to access the shared lun.

Configure an iSCSI target server and connect a client by performing discovery, scanning LUNs, logging in, and verifying shared storage with partitioning and systemctl-managed services.

Learn how to manage high-availability logical volumes with LVM and clustered LVM on shared storage, configure LVM and iSCSI targets, and understand active-active versus active-passive setups.

Discover how to share a disk across cluster nodes using iSCSI. Create a storage object from /dev/sdc, attach a lun, configure ACLs, verify wwn across nodes, and prep for lvm.

Configure a vm for linux high availability clustering using an iSCSI shared disk, create a physical volume, cluster vg, and a 500 MB cluster lv with zfs, enabling cross-node activation.

Configure lvm resources in a clustering environment by creating an lvm resource for a vg, then a zfs file system resource mounted at a mount point, validating failover across nodes.

Clustered LVM provides active-active access to shared storage across all nodes. It uses gfs2; ext4 and xfs are not supported, with VMD and DLM propagating metadata to prevent corruption.

Configure a clustered LVM in a three-node high-availability cluster using the distributed lock manager and CLVM, install LVM cluster packages, and create DLM and kvmd resources.

Create a volume group from a shared disk across nodes using a physical volume, then create a logical volume. Show that the vg and lv reflect on all nodes.