News, tips, partners, and perspectives for Oracle’s virtualization offerings

Oracle VM Server for SPARC 3.6.2 Released

Jeff Savit
Product Management Senior Manager

We are pleased to announce the release of Orcle VM Server for SPARC 3.6.2. This new release is integrated with Oracle Solaris 11.4 SRU 16 and provides bug fixes plus new and enhanced features. These include:

  • The add-vsan-dev command now supports live domain migration. One can now migrate a guest domain that has a vHBA instance associated with a virtual SAN with mask=on. This provides additional flexibility when virtual HBAs (vHBA) are used with devices selectively presented to domains.
  • New calculate-effective-max-pagesize option for add-domain and set-domain commands.  This option enables or disables the pagesize limit pre-determined by Oracle Solaris for a domain.

For further information, see the What's New page and the Release Notes on the Oracle VM Server for SPARC documentation page.  For additional resources:

Join the discussion

Comments ( 2 )
  • DavidH Friday, December 20, 2019
    Thanks for a great Christmas Gift!

    We would like to perform a cold migration of an LDom between the "cloud" and a datacenter, for D/R purposes, by having OVM for SPARC migrate storage differences!

    We would LOVE to perform a live migration of an LDom between the "cloud" and a datacenter, for D/R test, by having OVM for SPARC migrate storage differences!
  • Jeff Savit Friday, December 20, 2019
    Hi David,

    Thanks for the kind words!

    That's a great question, too. A lot depends on what your cloud system provides in terms of connectivity to your datacenter. Oracle VM Server for SPARC doesn't automate this, but if you have connectivity between the sites, you can copy virtual disk images between the two locations (limited by bandwidth, of course), and it's easy to get the guest domain constraints and use it to build the replica domain.

    For live migration it would be a bigger challenge, as you'd have to match networks and CPU types at both locations, and you would need common, shared storage across both.

    For DR the best approach would be to copy between different storage arrays and networks, to avoid a single point of failure of shared resources. Since OS images are large, it would be good to consider storage replication features such as those offered by ZFS appliances, or application-level replication, and try to send updates, "deltas" rather than complete instances.

    I hope that's helpful, and again - thanks for the nice comment. --- Jeff
Please enter your name.Please provide a valid email address.Please enter a comment.CAPTCHA challenge response provided was incorrect. Please try again.