Hi,

We've just performed our first upgrade to our standalone production FS7300, which went from 8.7.0.2 to 8.7.0.7. As this was a new'ish Flashsystem, it was installed as NPIV Enabled, and zoning in place according to all required hosts, of which the vast majority are VMware ESX hosts. We thought given NPIV we would see no errors in VMware for paths going offline/online, however we did see them this time. The time difference between path offline and path restored in VMware was only 10-15 seconds, but still, why did this occur, shouldn't NPIV avoid "all" path redundancy errors? 

The FS7300 is large, 16 x 38.4TB FCM4's, and has approx 30,000 - 40,000 IOps going through it when we did the upgrade, so it fairly busy. We are running 7.0.3 vSphere client.  

Is it still to be expected even with NPIV enabled that VMware may see drops of paths?

Cheers,

Andy

------------------------------
Andy Heath
------------------------------

Hi,

We've just performed our first upgrade to our standalone production FS7300, which went from 8.7.0.2 to 8.7.0.7. As this was a new'ish Flashsystem, it was installed as NPIV Enabled, and zoning in place according to all required hosts, of which the vast majority are VMware ESX hosts. We thought given NPIV we would see no errors in VMware for paths going offline/online, however we did see them this time. The time difference between path offline and path restored in VMware was only 10-15 seconds, but still, why did this occur, shouldn't NPIV avoid "all" path redundancy errors? 

The FS7300 is large, 16 x 38.4TB FCM4's, and has approx 30,000 - 40,000 IOps going through it when we did the upgrade, so it fairly busy. We are running 7.0.3 vSphere client.  

Is it still to be expected even with NPIV enabled that VMware may see drops of paths?

Cheers,

Andy

Yes, this is still expected,
The key prerequisite for smooth behavior is proper configuration;
Equivalent ports on both nodes must be connected to the same fabric. Hosts must be zoned to the NPIV WWPNs, not the physical ports. 

Hi,

We've just performed our first upgrade to our standalone production FS7300, which went from 8.7.0.2 to 8.7.0.7. As this was a new'ish Flashsystem, it was installed as NPIV Enabled, and zoning in place according to all required hosts, of which the vast majority are VMware ESX hosts. We thought given NPIV we would see no errors in VMware for paths going offline/online, however we did see them this time. The time difference between path offline and path restored in VMware was only 10-15 seconds, but still, why did this occur, shouldn't NPIV avoid "all" path redundancy errors? 

The FS7300 is large, 16 x 38.4TB FCM4's, and has approx 30,000 - 40,000 IOps going through it when we did the upgrade, so it fairly busy. We are running 7.0.3 vSphere client.  

Is it still to be expected even with NPIV enabled that VMware may see drops of paths?

Cheers,

Andy

Hi,

We've just performed our first upgrade to our standalone production FS7300, which went from 8.7.0.2 to 8.7.0.7. As this was a new'ish Flashsystem, it was installed as NPIV Enabled, and zoning in place according to all required hosts, of which the vast majority are VMware ESX hosts. We thought given NPIV we would see no errors in VMware for paths going offline/online, however we did see them this time. The time difference between path offline and path restored in VMware was only 10-15 seconds, but still, why did this occur, shouldn't NPIV avoid "all" path redundancy errors? 

The FS7300 is large, 16 x 38.4TB FCM4's, and has approx 30,000 - 40,000 IOps going through it when we did the upgrade, so it fairly busy. We are running 7.0.3 vSphere client.  

Is it still to be expected even with NPIV enabled that VMware may see drops of paths?

Cheers,

Andy