retrying all the portals again, since the portal list got exhausted

ESX has problems mounting iscsi LUNs and there are messages similar to this in the vmkernel log file:

Oct 15 04:01:37 xxxesx02 vmkernel: 11:00:26:19.095 cpu2:1078)<5>iSCSI: session 0x2e12010 iSCSI: session 0x2e12010 retrying all the portals again, since the portal list got exhausted
Oct 15 04:01:37 xxxesx02 vmkernel: 11:00:26:19.095 cpu2:1078)iSCSI: session 0x2e12010 to iqn.1998-11.com.netapp:sn.102503793 waiting 60 seconds before next login attempt
Oct 15 04:02:37 xxxesx02 vmkernel: 11:00:27:19.050 cpu1:1080)iSCSI: bus 0 target 2 trying to establish session 0x6e2c1b0 to portal 0, address 192.168.121.47 port 3260 group 1003
Oct 15 04:02:37 xxxesx02 vmkernel: 11:00:27:19.050 cpu1:1080)iSCSI: session 0x6e2c1b0 to iqn.1998-11.com.netapp:sn.102503793 failed to connect, rc -113, No route to host
Oct 15 04:02:37 xxxesx02 vmkernel: 11:00:27:19.050 cpu1:1080)iSCSI: session 0x6e2c1b0 connect failed at 116763511
Oct 15 04:02:37 xxxesx02 vmkernel: 11:00:27:19.050 cpu1:1080)<5>iSCSI: session 0x6e2c1b0 iSCSI: session 0x6e2c1b0 retrying all the portals again, since the portal list got exhausted

Solution

1. Copy the following files to /tmp:

/var/lib/iscsi/vmkbindings
/var/lib/iscsi/vmkdiscovery

2. Edit both of these files and remove names and ip addresses that are not in use

3. Rescan the iscsi hba

The error above will now go away and you will now either get access to this LUN or you'll get the error essage "may be snapshot: disabling access. See resignaturing section". See this article to solve that problem.

This issue has been observed after changing controller heads of a NetApp filer.