Computerworld - As Linux systems administrators watch over their hardware and software infrastructures, they constantly have to look ahead to how much space to allocate to hard-disk partitions to meet changing needs.
But the process of adding disks or reconfiguring existing drive space is slowed when servers have to be taken off-line for new drive installation and time-consuming rebooting or backup, repartitioning and data restoration.
Those changes can more easily be made using a logical volume manager (LVM), which precludes the need for reboots, shutdowns or downtime. Thus, an LVM can be an administrator's best friend.
Microsoft's Windows and Sun's Solaris and other Unix operating systems have similar volume manager capabilities.
Under Linux, an LVM is enabled by a kernel-loadable module, similar to a device driver in Windows. Kernel modules capture disk functions from user operations and convert them to write to the correct part of the disk. They work like a file system driver, but are far more dynamic.
Support for an LVM wasn't built into the Linux kernel until Version 2.4, but it was available as an add-on. It has been an important tool as Linux continues to make its way into enterprise computing.
"It's a common feature because it's so useful," says Chris Mason, a Linux kernel developer at Nuremburg, Germany-based Linux vendor SuSE Linux AG. "A lot of people, especially in Windows, may not know a lot about it," Mason says. "It's an extra layer to learn about, so a lot of people don't. Then when they move from a smaller server to a larger server, they don't realize how much more easily they can make it work."
The LVM creates an abstraction layer over all the combined storage in a system, so that the details about where the data actually resides are hidden. That allows the total separation of hardware and software because the LVM keeps a table of where the data is written and what volume group and volume it belongs to, allowing drives to be added or changed even while the system is running. And all this happens without software applications or users noticing the changes.
Systems without LVMs are also set up with physical hard drives, but they don't have this virtualization layer. When more disk space is needed, the system must be shut down and the new drives installed, and then they must be provided with file systems to organize data storage. If the new drive is replacing the original drive, the old drive must be backed up so the new drive can receive its data. That all takes time.
- 7 Elements of Radically Simple OS Migration OS migration is typically time-consuming and expensive. To make your next migration easy, follow these six recommendations when planning your project.
- Flying High on the Use of Red Hat Enterprise Linux Flybe was one of the 21 companies that were interviewed for quantitative results on their operations as part of an IDC ROI analysis....
- Who does NSS Labs "Recommend" for NGFW? In 2012, NSS Labs found that most available NGFW solutions "fell short in performance and security effectiveness." In 2013 NSS Labs noted "marked...
- 9 Essentials for a Complete Cloud-to-Cloud Backup Solution In 9 Essentials for a Complete Cloud-to-Cloud Backup Solution, we'll walk you through potential sources of data loss in the cloud and provide...
- Protecting Critical SaaS Data Before It's Too Late In this webinar, you'll hear how to avoid SaaS data loss through best practices from a panel of experts.
- What Does it Take to Deliver a Superior Customer Experience? The Two Top-Rated Online Retailers, B&H Photo and Crutchfield Electronics, Share Their Secrets Discuss practical CX tools and service methods such as contact center agents and the use of realtime speech analytics to help contact center... All Operating Systems White Papers | Webcasts