I have an EVA4100 and LUNs were presented to windows servers. Suddenly one server got hang and after reboot the server one LUN that presented to that server didn't appear. Finally I get it back after un present & present same LUN to the server.
I check the EVA from command view and couldn't find warning. But in the Controller A's host port connection status: Not logged into fabric Oparation status : Ok
What is the meaning of the "Not logged into fabric" and where can be the issue?
Fabric login is part of link intialization and happens after you connect a device into a switch. For some reason this did not happen for the controller. Apparently after re-presenting the LUNs the FLOGI did complete.
Unless someone was monkeying around with the cabling I would not suspect a bad cable, it would more likely be a bad SFP.
One way to figure it out is to telnet to the SAN switch (I am assuming Brocade) and type the command *porterrshow*. If there is more that 10% of the *enc out* errors than the total of the *rx + tx* then that would be a cable. If there are a bunch of *crc err* then that will point to a failed SFP.
Two things you could do quickly would be to: 1. From the telnet session disable and re-enable the port. *portdisable #* then *portenable #* 2. If you are using WWN zoning then move the cable to an open port with a populated SFP. If you are using Port zoning then swap the SFP with a known good one.
