Mirror management

BigFix provides tasks to help with the failback options for your mirror management solution. These tasks are available only for Solaris 10 endpoints.

System administrators create mirrors to protect data from disk failures. Disk mirroring, also known as RAID-1, provides high reliability and security against data loss. A mirror is composed of one or more RAID-0 volumes called submirrors. Each data block, which is a complete copy of each disk, is written to at least one other submirror. Typically, two or more disks are attached to a Solaris endpoint. In a scenario where a system has two disks, a mirror consists of two submirrors of the same data that is written on two different disks.

Each submirror contains an identical copy of the data root disk. When a disk fails, switching to one of the submirrors or disks can prevent system failure. Also, no data is lost because it can be recovered from other functioning submirrors or disks.

Note: BigFix does not support the creation or configuration of the root disk mirrors or submirrors. For more information about mirroring, see the Oracle documentation at http://docs.oracle.com/cd/E18752_01/html/816-4520/about-metadevices-24868.html.

Mirror management plays a key role in managing system downtime and risk when installing critical software patches or upgrades. BigFix provides the following tasks on the Patching Support site to help reduce system downtime and service interruptions due to patching:

Break Solaris mirrors: This task splits or breaks UFS file system mirrors, which in turn stops a mirrored disk from reading and writing to a backup disk.; For multiple file system mirrors on the root disk, the task will only break the mirror of the root file system.
Note: ZFS file system or VxVM based mirrors are not supported.

Re-mirror Solaris disks: This task resumes disk mirroring on Solaris 10 systems and synchronizes the updated disk with the backup disk.

As a precautionary measure before patching, use the Break Solaris mirrors task to split or break mirrors to allow you to use a disk as a backup copy of the system. You can then deploy the patch or upgrade to the master disk. You then run some verification steps to ensure that the patch or upgrade was successful. If the patch or upgrade was successful, you can use the Re-mirror Solaris disks task to put the disk back online, which begins the resynchronization process between the updated disk and the backup disk. However, if anything went wrong during the patching or the upgrade process, you can switch the disks and boot the backup disk without losing any data.