Comments on: Enterprise Computing: Migrating Petabyte Arrays http://thestoragearchitect.com/2009/01/19/enterprise-computing-migrating-petabyte-arrays/ Storage, Virtualisation & Cloud Mon, 21 May 2012 20:10:36 +0000 hourly 1 http://wordpress.org/?v=3.2.1 By: Chris Evans http://thestoragearchitect.com/2009/01/19/enterprise-computing-migrating-petabyte-arrays/#comment-542 Chris Evans Wed, 09 Feb 2011 12:53:44 +0000 http://thestoragearchitect.wordpress.com/2009/01/19/enterprise-computing-migrating-petabyte-arrays/#comment-542 Tushar Have these kind of appliances ever taken off? Firstly they have to be inserted into the data path, which can be a time consuming process and what about data integrity? One solution I think was most elegant was Softek's (now IBM) TDMF, but that relied on a host feature to make migrations work; of course on the mainframe there are only a handful of systems (or LPARs). Migration from large arrays could involve hundreds of hosts. Chris Tushar

Have these kind of appliances ever taken off? Firstly they have to be inserted into the data path, which can be a time consuming process and what about data integrity? One solution I think was most elegant was Softek’s (now IBM) TDMF, but that relied on a host feature to make migrations work; of course on the mainframe there are only a handful of systems (or LPARs). Migration from large arrays could involve hundreds of hosts.

Chris

]]> By: tushar http://thestoragearchitect.com/2009/01/19/enterprise-computing-migrating-petabyte-arrays/#comment-541 tushar Tue, 08 Feb 2011 11:56:20 +0000 http://thestoragearchitect.wordpress.com/2009/01/19/enterprise-computing-migrating-petabyte-arrays/#comment-541 Chris, Vendors like qlogic have come up with an appliance which sits right there in the storage path and continues online migration to heterogeneous arrays - so there is no downtime and heterogeneity is addressed and negligible performance impact. Chris,
Vendors like qlogic have come up with an appliance which sits right there in the storage path and continues online migration to heterogeneous arrays – so there is no downtime and heterogeneity is addressed and negligible performance impact.

]]> By: Chris Evans http://thestoragearchitect.com/2009/01/19/enterprise-computing-migrating-petabyte-arrays/#comment-540 Chris Evans Thu, 12 Feb 2009 10:08:21 +0000 http://thestoragearchitect.wordpress.com/2009/01/19/enterprise-computing-migrating-petabyte-arrays/#comment-540 Pete, great question - I have a number of answers to that - I'll put them together in a separate post. Pete, great question – I have a number of answers to that – I’ll put them together in a separate post.

]]> By: Pete Steege http://thestoragearchitect.com/2009/01/19/enterprise-computing-migrating-petabyte-arrays/#comment-539 Pete Steege Tue, 10 Feb 2009 16:40:16 +0000 http://thestoragearchitect.wordpress.com/2009/01/19/enterprise-computing-migrating-petabyte-arrays/#comment-539 Chris, If your scenario above were in a 100% virtualized server environment, how difficult would the migration be? What would it look like? Chris,
If your scenario above were in a 100% virtualized server environment, how difficult would the migration be? What would it look like?

]]> By: Chris "Saundo" Saunderson http://thestoragearchitect.com/2009/01/19/enterprise-computing-migrating-petabyte-arrays/#comment-538 Chris "Saundo" Saunderson Tue, 20 Jan 2009 22:45:00 +0000 http://thestoragearchitect.wordpress.com/2009/01/19/enterprise-computing-migrating-petabyte-arrays/#comment-538 I've got another for you:<br/><br/>1) Do currency!<br/><br/>Don't allow your volume managers, filesystems, operating systems and HBA/drivers to get too far out of date: support matrices are your friend on deciding what and when to replace the PB array with.<br/><br/>Working in a site that has 5+PB of disk, the biggest problem we run into is the lack of currency/patching of OS'/volume managers. Sneaky stuff like the number of PVLinks in a disk group can be worked around, but getting into the quagmire of upgrading volume managers kills any timeline, especially when conflicting support matrices end up forcing OS upgrades along with the VM or HBA upgrades! I’ve got another for you:

1) Do currency!

Don’t allow your volume managers, filesystems, operating systems and HBA/drivers to get too far out of date: support matrices are your friend on deciding what and when to replace the PB array with.

Working in a site that has 5+PB of disk, the biggest problem we run into is the lack of currency/patching of OS’/volume managers. Sneaky stuff like the number of PVLinks in a disk group can be worked around, but getting into the quagmire of upgrading volume managers kills any timeline, especially when conflicting support matrices end up forcing OS upgrades along with the VM or HBA upgrades!

]]>