I did come across some behavior I hadn\u2019t witnessed previously.\u00a0 We were dealing with a CX4-120 that had been IPed on a NATed network; to include hba initiators that had been presented from ESXi hosts on a local domain, versus the lab domain already in existence.<\/p>\n
Some might say it was an unintended \u201cair-gap\u201d solution, I call it an island in the Bermuda Triangle, figure that one out Breitling lovers.\u00a0 The tasking at hand was to re-provision an entire Vblock0, which really was almost a Vblock1 because it had been upgraded to the FC back-end.\u00a0 Nonetheless, it needed new IP\u2019s, an herbal remedy on the array scenario (Flare code update, zoning, re-ip, RAID\/Thin Pool\/LUN reconfigure), reconfigured network, install UIM and scale-out a Citrix XenDesktop deployment\u2026in 3 days, short order for 4 technicians.<\/p>\n
It was a good thing Duane Schweitzer (VCE) was on hand.\u00a0 We sat and scratched our heads for a few hours regarding the seemingly anomalous behavior that was occurring.\u00a0 As we see below, the host connectivity status shows a HBA WWPN presented, without the hostname. \u00a0We see the new hosts presented (vblade*.lab.*.*.*) and we see the old hosts from the previous domain (*.vblock.local) with registered initiators.\u00a0 Note that the entire environment was a mixture of initiators and the old hosts with the initiators still registered.\u00a0 The below screenshot is near the end of our reconfigure.<\/p>\n
What would you do?\u00a0 Look on Powerlink for the Unisphere Administration Guide?\u00a0\u00a0 Wrong.\u00a0 Contact a CX4 guru?\u00a0 Wrong.\u00a0 Call Watson?\u00a0 It was busy destroying the mortals.\u00a0 Thus, the smartest thing in the world couldn\u2019t help and it has AT&T service, so we knew the celly was about useful as salt while floating in the middle of the ocean.\u00a0\u00a0 It seems relatively obvious what the fix action is, but not until we arrived to it.<\/p>\n In our discussions, we knew that the initiators had been registered with ESXi hosts under the 192.168 address space.\u00a0 How would we know which initiators were which?\u00a0 Good thing it\u2019s a VBlock, with UCS.\u00a0 As you may or may not know, the HBA WWPNs are part of the UCS object repository or as they refer to them\u2026vHBAs.\u00a0 Comparing the Unisphere Connectivity Status window with the specific vHBA on the host, we see 2 registered with the old host and two registered, but not associated with Host 12 (vXXXXXX.vblock.local).<\/p>\n Now we have a starting point.<\/p>\n Right click the highlighted WWPN and select edit.<\/p>\n The \u201cSelected Host\u201d radio button will be selected.\u00a0 Change the radio button to the \u201cNew Host\u201d.\u00a0 Provide the new Host Name and the correct IP Address.\u00a0 I can\u2019t explain the Initiator Type, but recalled reading somewhere it needed to be CLARiiON Open and I knew using ESXi Boot from SAN required failovermode 4.\u00a0 This will assign the correct host and IP to the \u201crandom\u201d WWPN.\u00a0\u00a0 In addition, this will add the new host and IP to the \u201cBrowse Host\u201d section that we\u2019ll see down below.<\/p>\n Returning to the connectivity status, right click and select \u201cedit\u201d on the host in question.\u00a0 In this example it was vXXXXXX.vblock.local.\u00a0 Like all techs, screenshots are a consortium of free flowing thoughts; use your imagination on this one.<\/p>\n Returning to the screen we\u2019ve seen before in this write-up, this time select the \u201cExisting Host\u201d radio button and browse.<\/p>\n![\"\"](\"http:\/\/www.jasemccarty.com\/blog\/wp-content\/uploads\/2011\/02\/tc01.png\" "\\"tc01\\"")
<\/a><\/p>\n<\/a><\/p>\n<\/a><\/p>\n<\/a><\/p>\n<\/a><\/p>\n