toasters August 2015

toasters@lists.teaparty.net

24 participants
16 discussions

FAS2554 \ ADP
by Mike Gossett 06 Aug '15

06 Aug '15

Hi all, First, things for a great forum. I learn a lot from reading these posts, and I appreciate everyone sharing their experiences. I am in the middle of a transition for a client whom had 2 2040 systems (HA pair at primary site; single controller at DR) to an 8020 c-mode system at HQ, and a FAS2554 at DR. We are using hybrid sata\flash pool at the primary, and will have 3 total shelves (including the onboard shelf in the 2554) of 2TB sata out of the gate. Typically, I am deploying smaller systems with 24 disks where the ADP is highly desirable from a performance and capacity standpoint. However, in this instance, I am leaning towards converting the system over to using dedicated cluster disks, such that the balance of the disks onboard the FAS2554 are the same size as the other two shelves, so I can have 1-2 large aggregates, and not be limited by the boundary of the artificially smaller shared ADP disks. So, I really have two questions. First, what have your experiences been? Would you agree with my gut feeling that I am better off having the dedicated cluster disks? Capacity in this environment is growing somewhat rapidly, and I would expect that in 2 years we will be adding at least one more shelf to the mix. Second, assuming you feel that moving off ADP is the right approach, can anyone point me in the direction of an article or high level steps in terms of how to accomplish this? I'm fine with disruptive or destructive activities as the system is not yet in DR (though I've completed the transition at HQ to the new cluster and decommed the legacy 7m system, and as such am anxious to resume replication). Thanks in advance, Mike

1 0

Re: CIFS on Mac
by Ryan Pugatch 06 Aug '15

06 Aug '15

Just wanted to follow up - does anyone have shares with significant amounts of data and experience slow directory listings when accessing from Mac via CIFS? Thanks, Ryan On Tue, Aug 4, 2015, at 05:24 PM, Ryan Pugatch wrote: > Actually, to Andre's point, spotlight doesn't work on CIFS, only AFP. > > In talking with one of our desktop team folks, the specific issue > isn't due to indexing (though that is an issue, not having it..) but > specifically getting directory listings on the CIFS share from OS X is > very slow. Windows users do not have that problem. > > > On Tue, Aug 4, 2015, at 05:22 PM, Tony Bar wrote: >> Ryan – >> >> You can disable spotlight on a per-filesystem basis, see the bottom >> of the link for details. >> >> *Anthony Bar* >> 650.207.5368 >> tbar(a)berkcom.com >> *Berkeley Communications* >> www.berkcom.com[1] >> >> *From:* Ryan Pugatch [mailto:rpug@lp0.org] >> >> *Sent:* Tuesday, August 4, 2015 2:19 PM *To:* Tony Bar *Cc:* >> Toasters(a)teaparty.net *Subject:* Re: CIFS on Mac >> >> I don't think we would want to disable spotlight across the board. >> It is useful in some cases. >> >> Thanks, >> Ryan >> >> >> On Tue, Aug 4, 2015, at 04:03 PM, Tony Bar wrote: >>> Ryan - >>> >>> This may help: >>> >>> http://recomhub.com/blog/how-to-disable-or-enable-spotlight-in-mac-os-x-yos… >>> >>> Anthony Bar >>> 650.207.5368 >>> >>> tbar(a)berkcom.com Berkeley Communications >>> >>> www.berkcom.com[2] >>> >>> On Aug 4, 2015, at 1:01 PM, Ryan Pugatch <rpug(a)lp0.org> wrote: >>> >>>> Hi all, >>>> >>>> Performance when accessing netapp CIFS shares from a Mac is >>>> frustrating, >>>> and in some cases downright unusable. >>>> >>>> It seems mostly related to OS X's desire to index everything with >>>> Spotlight. Directory listings are slow, and if you're any way >>>> distant >>>> from the netapp, you'll find latency makes it completely unusable. >>>> >>>> Have any of you found how to deal with this? >>>> >>>> Thanks, >>>> Ryan >>>> _______________________________________________ >>>> Toasters mailing list >>>> Toasters(a)teaparty.net >>>> http://www.teaparty.net/mailman/listinfo/toasters > > _________________________________________________ > Toasters mailing list Toasters(a)teaparty.net > http://www.teaparty.net/mailman/listinfo/toasters Links: 1. http://www.berkcom.com/ 2. http://www.berkcom.com/

1 0

Virtualized Exchange 2010 on cDOT
by Milazzo Giacomo 05 Aug '15

05 Aug '15

Hi all, we're going to migrate a running environment in 7Mode to cDOT and, nevertheless I've studied all docs about best practice and DR scenarios I've a lot of doubts about best architecture. Now we have. Production site. 2 E2K10 virtualized Server with a two copies of db in DAG environment These 2 servers have their data on different controllers of an HA system and db are balanced on two controllers/server some as active and some as passive (but you can imagine also one server running active copies and the second the passive ones) Servers are this way built: C:\ system disk with Windows and Exchange binaries on a NFS vmdk with vmx in the same path Database disks on VMware iSCSI raw devices (this because Microsoft does not support NFS vmdk for its data...no comment!..we know...) For obvious reason there's also another little iSCSI datastore containing the "pointer" to those raw device, the mapper (rdm). This is the way it works Raw devices are managed by SnapDrive for there's also SME running for backup/mirror. DR site At recover site there's an identical enviroment with all the volumes replicated using SnapMirror modulated by SME VMware SRM together with SRA mount, in case of disaster, the right vmdk, LUNs and so on using its own script so the E2K10 servers at DR can start with all DAG One of the constrain of SRM/SRA in 7Mode is that ALL the volumes with NFS vmdk, iSCSI raw devices and iSCSI ds with rdm MUST belong to the same controller; so that is not possible to spread volumes on different controller/aggregates maybe to increase resiliency...Otherwise SRA will fail. What's about cDOT? Can I create two different SVM, one just for NFS vmdk for binaries and the second one for iSCSI contents? Or have/may have I to put all togheter? Have I to create, as stated in DR scenarios TR from NetApp, two SRM to manage the two Exchange servers or, maybe, to relocate data on different aggregates? Have these two aggregates to be resident on the same node or they can belong to the same node? What's about SRM/SRA constrain? Thanks a lot for any suggestion/idea. Regards,.

2 2

CIFS on Mac
by Ryan Pugatch 04 Aug '15

04 Aug '15

Hi all, Performance when accessing netapp CIFS shares from a Mac is frustrating, and in some cases downright unusable. It seems mostly related to OS X's desire to index everything with Spotlight. Directory listings are slow, and if you're any way distant from the netapp, you'll find latency makes it completely unusable. Have any of you found how to deal with this? Thanks, Ryan

7 10

Re: CIFS on Mac
by Ryan Pugatch 04 Aug '15

04 Aug '15

Yeah, we tried that - no difference. Ryan On Tue, Aug 4, 2015, at 04:26 PM, Clark, André M. wrote: > Spotlight is only supported on AFP-based network volumes. CIFS and > NFS volumes aren’t indexed. > > Are you connecting to those shares via cifs://server/share or > smb://server/share? There was an issue with performance if you were > mounting via cifs:// vs. smb://. > > Regards, > > André M. Clark[1]| Sr. Solutions Architect | 917.388.8236 *Tell me I > will forget... Show me I may remember... Involve me I WILL > UNDERSTAND!!! > Start by doing what's necessary, then what's possible, and suddenly you are doing the impossible!!!* > *Red8*[2], An Insight Investments Company > www.red8.com[3] > > *From:* toasters-bounces(a)teaparty.net [mailto:toasters- > bounces(a)teaparty.net] *On Behalf Of *Tony Bar *Sent:* Tuesday, 4 > August, 2015 16:04 *To:* Ryan Pugatch <rpug(a)lp0.org> *Cc:* > Toasters(a)teaparty.net *Subject:* Re: CIFS on Mac > > Ryan - > > This may help: > > http://recomhub.com/blog/how-to-disable-or-enable-spotlight-in-mac-os-x-yos… > > Anthony Bar > 650.207.5368 > tbar(a)berkcom.com > Berkeley Communications > www.berkcom.com[4] > > > On Aug 4, 2015, at 1:01 PM, Ryan Pugatch <rpug(a)lp0.org> wrote: >> Hi all, >> >> Performance when accessing netapp CIFS shares from a Mac is frustrating, >> and in some cases downright unusable. >> >> It seems mostly related to OS X's desire to index everything with >> Spotlight. Directory listings are slow, and if you're any way distant >> from the netapp, you'll find latency makes it completely unusable. >> >> Have any of you found how to deal with this? >> >> Thanks, >> Ryan >> _______________________________________________ >> Toasters mailing list >> Toasters(a)teaparty.net >> http://www.teaparty.net/mailman/listinfo/toasters Links: 1. mailto:Andre.Clark@redeight.com 2. http://www.red8.com/ 3. http://www.red8.com/ 4. http://www.berkcom.com/

1 0

stuck on a configuation puzzle: NFSv3 in 8.3, netgroups and export rules
by Rue, Randy 04 Aug '15

04 Aug '15

Hello All, Sent the below to our NetApp contacts but often you beautiful people have the answers I need, better and before the vendor does. Need to limit an NFS export on an 8.3 SVM to either a NIS netgroup or an AD/LDAP nisNetgroup, preferably the LDAP. See the below. Since I wrote the below, I've also determined that the working vserver appears to have a working copy of the netgroups in cache and the broken one does not: gold::*> vserver export-policy netgroup cache show -vserver foo There are no entries matching your query. gold::*> vserver export-policy netgroup cache show -vserver pyrite-gold Vserver Netgroup State -------- ---------- ------------ pyrite-gold netgroup-cluster-rcs.fhcrc.org ready The documentation says that "One of the methods you can use to match clients in export policy rules is by using hosts listed in netgroups. You can load netgroups from a uniform resource identifier (URI) into Storage Virtual Machines (SVMs) as an alternative to using netgroups stored in external name servers (vserver services name-service netgroup load)." I'm specifically hung up on "as an alternative to using netgroups stored in external name servers." What is the method I'm missing that manually loading the netgroups via a URL (yuck) is an alternative to? Hope to hear from you, Randy in Seattle (sent to NetApp) ----------------------------------------------------------- Guys, We're attempting to set up NFS access for an SVM to a NIS netgroup or its LDAP equivalent and while I was able to do this once on a test SVM we're now unable to recreate it. On our old gear, we had some issues trying to make NIS work and eventually scripted a flat file dump of our netgroup file to the vFilers, then used the local file for hostgroup based access to NFS exports. That group name is @CLUSTER in /etc/exports That same script also takes the members of our netgroup file and exports them to an LDAP object in our AD, a nisNetgroup object called netgroup-cluster-rcs.xxx.org. That group is currently added to the export policy as @netgroup-cluster-rcs.xxx.org Listing the AD object as above works for an SVM called pyrite-gold, a test vserver. My recollection is I set that one up pretty easily after realizing NIS wasn't working. I don't recall doing anything at the CLI. I can mount an export quickly and correctly from a linux host that is a member of the group. Now we're trying to set up the same config on other SVMs ("foo") and document the steps for production and can't make it work. We've compared the config on the new SVM and pyrite-gold every which way with no luck. Gone over the GUI config a handful of times. I've also looked deeper from the CLI and found a few differences but changing them didn't solve the problem. When we try to mount the export, the attempt retries a handful of times and eventually times out: [root@RANDINO mnt]# mount foo:foo_nfs foo/ mount: mount to NFS server 'foo' failed: timed out (retrying). mount: mount to NFS server 'foo' failed: timed out (retrying). mount: mount to NFS server 'foo' failed: timed out (retrying). mount: mount to NFS server 'foo' failed: timed out (retrying). mount: mount to NFS server 'foo' failed: timed out (giving up). [root@RANDINO mnt]# >From the CLI, I did find that ns-switch for "foo" was set only for "files." Fixed that using vserver services modify with no improvement. Now all matches: gold::*> vserver services name-service ns-switch show Source Vserver Database Order --------------- ------------ --------- foo hosts files, dns foo group files foo passwd files foo netgroup ldap, nis, files foo namemap files pyrite-gold hosts files, dns pyrite-gold group files pyrite-gold passwd files pyrite-gold netgroup ldap, nis, files pyrite-gold namemap files I also took a look at LDAP client settings: gold::*> vserver services ldap client show -instance Vserver: foo Client Configuration Name: foo LDAP Server List: xxx Active Directory Domain: fhcrc.org Preferred Active Directory Servers: xxx Bind Using the Vserver's CIFS Credentials: true Schema Template: RFC-2307 LDAP Server Port: 389 Query Timeout (sec): 3 Minimum Bind Authentication Level: anonymous Bind DN (User): sops-ldap Base DN: dc-xxx,dc=org Base Search Scope: subtree User DN: - User Search Scope: subtree Group DN: - Group Search Scope: subtree Netgroup DN: - Netgroup Search Scope: subtree Vserver Owns Configuration: true Use start-tls Over LDAP Connections: false Allow SSL for the TLS Handshake Protocol: false Enable Netgroup-By-Host Lookup: false Netgroup-By-Host DN: - Netgroup-By-Host Scope: subtree Vserver: pyrite-gold Client Configuration Name: pyrite-gold LDAP Server List: xxx Active Directory Domain: fhcrc.org Preferred Active Directory Servers: xxx Bind Using the Vserver's CIFS Credentials: true Schema Template: RFC-2307 LDAP Server Port: 389 Query Timeout (sec): 3 Minimum Bind Authentication Level: anonymous Bind DN (User): tinsel-ldap Base DN: dc=xxx,dc=org Base Search Scope: subtree User DN: - User Search Scope: subtree Group DN: - Group Search Scope: subtree Netgroup DN: - Netgroup Search Scope: subtree Vserver Owns Configuration: true Use start-tls Over LDAP Connections: false Allow SSL for the TLS Handshake Protocol: false Enable Netgroup-By-Host Lookup: false Netgroup-By-Host DN: - Netgroup-By-Host Scope: subtree 2 entries were displayed. gold::*> Note that "Bind Using the Vserver's CIFS Credentials" was set to false for foo, but again, changing that from the CLI didn't make any difference. I know I never touched that setting on pyrite-gold, is it set that way because (maybe) I ran or re-ran CIFS setup after setting up NIS? We need help on this pretty quickly: either a way to make NIS work for netgroup control of an NFS export policy, or a way to make it work based on the AD nisNetgroup. Given a choice we'd prefer the LDAP way. Is there a NetApp doc I've overlooked? Randy

4 3

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

1998

1997

toasters August 2015