It's not currently supported to mix the speeds. Page 38 of the 8.0 FC/iSCSI config guide does document that for the FAS3200: ------ Each 32xx controller supports 4-Gb and 8-Gb FC target expansion adapters. Note that the 8-Gb expansion adapters cannot be combined with 4-Gb targets (whether using expansion adapters or onboard). ------ Regarding he BSAG comment: ----- Considerations If you install target expansion adapters on the system, you must configure the onboard adapters as initiators. ----- This was a left over from when it was not allowed to mix onboard and add-in expansion target ports. I've marked that and we'll work on getting that straighted out. We'll also work on getting the statements on the mixing of speeds to be more definitive. Cheers, Rick
... View more
Hi, While I haven't tested this specifically I do know that the SVC does NOT zero the LUNs so I would expect there to be not issue. Cheers, Rick
... View more
There aren't really any downsides to letting SnapDrive do it automatically. The ability to do manual igroup management was added for people that want to be able to specify particular igroup names, etc... . The unique LUN IDs is not really a SnapDrive "issue" but storage system topic. On an active/active (or clustered) storage system, the LUNs IDs do have to be unique since the LUNs are available across the ports on both controllers with the same WWNodeName across controllers. When mapping LUNs the storage system checks if the number (the N in LUN) is already used for that igroup and only allows unused numbers. Cheers, Rick
... View more
The DSM is not included as part of the host utilities. If you have an active/active system with connections to both controllers you'll have mulitple paths to the LUN even if you only have a single path in the host (assuming the host can access both controllers). With the HU and DSM I'll bet all will be well.
... View more
Hi, I don't believe we have any best practices specific to ORACLE ASM on Vmware and iSCSI. I would generally recommend RDMs in combination with ASM to have more flexibility in terms of move data between system types and I don't see the advantages that VMFS would provide you in this environment. RDM's do have the "extra" management overhead of needing to map multiple LUNs. As for the sizeof the LUNs, it depends on the size of the database. For max performance I'd generally try to some reasonable small number of LUNs. From the NetApp point of view the number of LUNs doesn't really affect performance but that's not true will all OSs so I generally recommend multiple LUNs (4-8) for an application since it still realtively easy to manage but still within the sweet spot of where most configuration are running. Cheers, Rick
... View more
Hi, There's no hidden snapshot but when the volume is dedup'd and there are space reserved objects (typically LUNs are the only space reserved object but it also possible to enable space reservations on files) the space will be reserved. If you disable reservations on the LUNs, which generally makes sense when using dedup, you should see the space returned. Also, as mentioned in this case the volume guarantee == none and because of that even though the space is "reserved" at the volume level no physical blocks at the aggregate level are being reserved.
... View more
> vol options: Volume option fractional_reserve can be modified only if the volume has its guarantee option set to 'volume'. > Is there a way to configure to configure a volume, so the free block will be returnm to the aggr with fractional reserve = 0? Currently, the fractional_reserve can only be set to 100% when guarantee=none. This will be changing in a future ONTAP release so that we can have guarantee=none and fractional_reserve=0%. As Sajan mentioned with a guarantee=none the fractional_reserve doesn't actual reserve physical space but we want to allow setting it to 0% in all cases since that is now our general best practice and it's easier from a managability aspect.
... View more
I'm not aware of any numbers that we NetBackup numbers but Romeo seems to have some pretty good numbers and someone else might be aware of some NetApp numbers. Deduplication is really protocol independent so we wouldn't expect different numbers based on protocol. Cheers, Rick
... View more
You certainly can and will want to use them together. Our best pratices paper on VMware cover this in very good detail. The link to those Best Practices is: http://media.netapp.com/documents/tr-3428.pdf I think the info that you're looking for starts on pages 53 and 56. Let us know that works. Cheers, Rick
... View more
> Where do these limits come from and why do they vary from system to system? For example, why is the limit 4TB on the FAS 3140 and 16TB on the FAS 3170? The limits come from the available resources on the systems (CPU, memory, etc...). > Is there any work-around for this? The limits are volume based so one can break things into multiple volumes. That obviously has some trade-offs in terms of deduplication as well as other areas but is certainly a possibility.
... View more
> I have a 280GB FlexVol with a single LUN of 140GB. It > currently has 87GB of data on it.The volume Snapshot Reserve > is set to 0 and the LUN has Fraction Reserve set at 100. > > A df -rh reveals that the volume is using 259GB....there is > only 2.9GB of snapshots. How to I determine what is using all > the space? The df -rh should show that ~116GB is reserved. The LUN is taking 140GB, the snapshots 2.9GB and the reserve is 116GB. My assumption is that the host fileystem is showing 87GB used but there is actually 116GB of data in the LUN and therefore the reserve is also 116GB. TR3483 goes into gory detail regarding this. > I have RTFM'd (Read the Fabulous Manual) the Block Mgt Guide; > there is a great section that explains Fractional Reserve, > etc-but I just not "getting it". > > Perhaps there is something on NOW that explains this better > than the Block Mgt Guide? Maybe Blocks for Dummies? With 7.2.4 we are in the process of changing all our documentation to the standard recommend of setting fractional reserve = 0 and enabling autodelete (trigger = volume) to be sure that there is always space available. I refer to this as the "X + Delta" configuration. This is not only much more efficient it is infinitely easier to understand because the space consumed is equal to the LUN space (whatever size the LUN or LUNs are configured as) and the snapshot space: Used Space = LUN(s) size + Snapshot space Cheers, Rick
... View more
The igroup information is storage internally and is available through the CLI (igroup command) and the API calls. There is no other supported or documented method to access this information. > Is it also true, that when a head is upgrade from a 3020 to 3040 Unfortunately this was true but has been fixed with the 7.2.4 release. Starting with 7.2.4 the serial number will not change when the controller is updated. > I've noticed that the LUN serial numbers are different on our SnapMirror copies at our DR site, ..... > Why do LUN serial numbers change? This is by design to provide data integrety. When ever LUNs change "locale" (volume, storage system, etc...) the serial numbers are changed. There is currently no method to retain the original serial numbers. It is possible to set the serial #'s to the original values using the "lun serial" command or api.
... View more
