# zpool online -e rpool /dev/dsk/c7d0█
New and Improved. It’s a term marketers use all the time to try to lure us in either to purchase a product or support a change. While I haven’t seen Oracle advertise the revamp of My Oracle Support (MOS) as “new and improved,” what I can tell you is that it is “new and destroyed.”
Continue reading “Oracle’s MOS revamp fails spectacularly”Older NFS servers don’t always talk to newer clients. Here’s a workaround until you can update the server. Continue reading Modern NFSv4 clients with Solaris 10 servers
Most operating systems make certain assumptions about environments in order to make configuring them “easier.” Those assumptions are usually based on what the majority of users need or on what the industry thinks your environment should look like. The NFSv4 domain name in Solaris is a good example of one of those assumptions that is … Continue reading Using svccfg to set the NFSv4 domain name
Hard drive firmware updates used to be a scary process that involved downtime and required booting from an alternate source such as DVD. Solaris 10 allowed us to have mirrored root zpools starting with the 10/08 (u6) release, which allows us in many cases to update drive firmware essentially non-disruptively. First of all, decide which … Continue reading Solaris non-disruptive drive firmware updates
While it is possible to run a zpool on thin-provisioned storage, not all versions and patch levels of Solaris support the Logical Block Provisioning (LBP) SCSI feature. There’s even some question as to whether the versions that do support LBP support it well enough. It’s not wise to thin provision the underlying storage unless you’re … Continue reading Thin provisioning storage for zpools
Many modern SAN storage systems have a built-in snapshot capability. There are some valid use cases for such a capability but generally speaking, you’re not likely to run into those cases for volumes that are hosting LUNs for advanced filesystems like ZFS. Many of these advanced filesystems assume they have full control over the devices they’re managing and they therefore don’t handle out-of-space conditions on those devices well. One example is below.
I ran into a situation where I needed to duplicate a Solaris 11 server from the primary data center to a DR site 850 miles away. A couple firewalls and a WAN separate the source from the target. I’m sure there are other ways to accomplish a feat like this but I will describe here the way I decided to do it and a little bit about why.
Continue reading “Duplicate a root pool across multiple firewalls”