Quantcast
Channel: ONTAP Discussions topics
Viewing all 4994 articles
Browse latest View live
↧

Node Root Volume Rename

August 16, 2016, 10:38 am
$
0
0

Hey everyone.  First post to the community boards.  IHAC that changes the root volume names to align with their internal nomenclature for their systems.  In testing out ONTAP 9, it looks like this capability is not there.  Here's the errors I get:

 

svl2-affselab01::volume> rename -vserver svl2-affselab01h01a -volume svl2_affselab01h01a_root -newname vol0 -foreground true

Error: command failed: Internal Error. Failed to determine the SnapLock type of the volume "svl2_affselab01h01a_root" in Vserver "svl2-affselab01h01a". Reason: entry doesn't exist.

 

If I try using the cluster name as the vserver, here's what I get:

 

svl2-affselab01::volume> rename -vserver svl2-affselab01 -volume svl2_affselab01h01a_root -newname vol0

Error: command failed: This operation is not supported for the system volume "svl2_affselab01h01a_root".

 

I would appreciate any help with this.  Thanks!

↧
Search

SMB2 QFid Support

August 16, 2016, 6:36 pm
$
0
0

Hi:

 

I have a customer who has contacted Microsoft in regards to some performance issues.  MSFT have come back asking if their CIFS storage supports SMB2 Qfid.  As far as I can tell, it references the following article:

 

https://msdn.microsoft.com/en-us/library/cc246504.aspx

 

MB2_CREATE_QUERY_ON_DISK_ID

0x51466964

("QFid")

The client is requesting that the server return a 32-byte opaque BLOB that uniquely identifies the file being opened on disk. No data is passed to the server by the client.

 

The customer is connecting to NetApp storage (clustered OnTap 8.2.3P6).  I am unable to find any references to this anywhere in documentation or the NetApp support site.  The only reference I could find was in the file access guide for 7-mode version 7.x in which it simply states it is not supported as they could find no reason for using it.  Anyone here have any knowledge of this?  Thanks.

 

-Mike-

↧
↧

NFS mounting vol0 of Cluster Node

August 17, 2016, 1:45 pm
$
0
0

In cDOT enviromnet, Is there any way to NFS mount an Vol0 of node on a Linux machine..?

 

↧

creation snapshot of aync option

August 18, 2016, 4:29 am
$
0
0

During the test of snapshot creation of Flex volume, I've found async option in "snapshot-create" API.
As far as I know, there is no async option in ONTAP8.3.1P1 "snapshot create" command.
So how can I know whether this async option working?
Is there some kind of status which shows like "Snapshot creating..." or "Snapshot created" ?

 

<?xml version="1.0" encoding="UTF-8"?>
<netapp  xmlns="http://www.netapp.com/filer/admin" version="1.31">
  <snapshot-create>
    <async>true</async> ★
    <snapshot>test_snap
    </snapshot>
    <volume>volume_iscsi</volume>
  </snapshot-create>
</netapp>

↧

Adding shelf, Disk ownership, assignment and auto aggregate creation?

August 18, 2016, 8:30 am
$
0
0

Trying to add a new shelf of disks to existing cluster.  I've got an issue where I can assign the new shelf of disks to a node with no problems but after assigning the shelf or even the disks one at a time it automatically creates an aggregate (aggr0,aggr1, aggr1(1) ) with those disks. Which sounds good right,  but the aggregate seems to be an aggregate that I can't do anything with. What makes it worse is that CDOT (8.1) doesn't seem to even recognize that aggregate that was automatically created as a true aggregate because It won't even show in the list of aggregates when I run aggregate show. If this were the case I should be able to delete the aggregate and those disks should then show as spares. Basically what I would like to do is simply assign ownership of that shelf to a node and have these disks show up as hot spares so that I can actually do something with the disks (like create an aggregate then from there create volumes/LUNs on that aggregate). This is the way it should work under normal circumstances. Not the case here.

Has anybody ever ran into this issue? Any help would be much appreciated!

↧
↧

Efficient method for deleting a large amount of data

August 18, 2016, 11:45 am
$
0
0

Hello,

 

We're running OnTap 8.2.1. I was looking for a system side command I could use to browse cifs shares and remove a particular folder under a root share as opposed to navigating the share from Windows, but I haven't seen anything. I'm curious if anyone here has a tip/trick for removing overly complex, deep/large folder structures without just removing the entire share?

 

So far it's taking 20+ minutes to discover the folders content from a windows client before the system starts to delete anything.

 

The structure looks similar to the following:

 

Groups (top level folder/share)

-Group1 (1TB, 250k+ files)

-Group2 (2TB 1M+files)

-Group3 ( 1.8TB 500k+ files)

 

The "group 2" folder would be the one I'd like to remove.

 

Appreciate any assistance you could provide.

 

-Brian

 

 

↧

Login to different igroup

August 18, 2016, 7:51 pm
$
0
0

Hi Team,

 

I am very new to storage technologies

 

we have FAS3140 and NetApp Release 8.0.4 7-Mode It has two controllers.

 

I have created Lun, Iguroup and mapped lun to created igroup

 

igroup1_I (iSCSI) (ostype: linux):
        iqn.2000-01.com.linux:build:1616 (not logged in)

 

/vol/boot/lun212              20g (21474836480)   (r/w, online, mapped)

 

when i tried login from linux initiator to above igroup it is log in to different i group that is in another controller

 

filer-4> iscsi show initiator
iscsi show initiator: This command has been deprecated.
Please consider using 'iscsi initiator show' instead.

Initiators connected:
  TSIH  TPGroup  Initiator/ISID/IGroup
   350    1004   simha2.blr.asicdesingers.com (iqn.2003-11.com.chelsio:vidya:1234 / 00:02:3d:04:00:00 / )

 

filer-4> iscsi nodename
iSCSI target nodename: iqn.1992-08.com.netapp:sn.151762993

 

Please help us to fix this issue. I dont know if you are understand my problem in case not please let me know what detail i need to provide.

 

 

 

 

↧

Is there a list of production safe commands on the node shell?

August 19, 2016, 12:11 am
$
0
0

Hello NetApp Community, fellow admins and technicians,

 

I had a small and simple performance analysis script on my 7-mode systems that I was using every once in a while to see the current load of a specific 7-mode system. The script was as follows:

-----8<--------

priv set -q diag
nfsstat -z
statit -b
sysstat -x -s -c 60 2
echo ""
echo "statit -e"
echo "========="
statit -e
echo ""
echo "nfsstat -d"
echo "=========="
nfsstat -d
priv set

-----8<--------

 

I would run it via the "source" command on my filers CLI. I was happy with it.

 

*Enter Clustered Data ONTAP*

 

Well, I was trying to get some performance information out of CDOT and let's put it politely: CDOT is not talkative. Looking for solutions I stumbled over the OnCommand Performance thingy and it's VM requirements are simply insane. NetApp, I'm not gonna waste so many cores and cycles and memory on our VMWare cluster just to get only a subset of possibilities that I had with DFM. ... Well and of course then I still would have to run DFM in parallel because 7-mode.

 

...

 

Now I know that there are commands that can be savely run on a cluster node, eg. sysstat. Other commands are not supported on the node shell anymore and might actually panic your system, let me put it straight: This behaviour is acceptable, but my boss would still tear my head off if I make our systems panic.

 

Now to the point: I was wondering, is there an official list of commands that are known to be safe (or dangerous) to be ran on the node shell?

 

Best regards,

 

Konstantin

 

 

↧
Search

copy data with non-utf8 file names to nfs4 volume

August 19, 2016, 6:37 am
$
0
0

We've got a linux file system (ocfs2) with file names encoded with iso 8859-1. We'd like to copy this data into a a new nfs4 share from a 7-mode filer. The share is mounted on the same linux server. In an iso 8859 environment trying with cp, scp or tar-copy-untar this fails when the data is to be written to the share:

cp -p /source_path/* /target_path/
cp: cannot stat `./fleißige_Azubis.jpg': Invalid argument

 

Obviously this is due to the German special character encoded with a single byte xDF. We have been trying with different language configuration of the volumes shared as NFS4 but neither of the languages (en_US, en_US.UTF-8, de, de.UTF-8) worked.

 

Is there a way to get this data copied? - Ideally by maintaining the encoding as there is a database containing rerferences to the file names.

 

 

 

 

 

↧
↧

Is lun alignment show command removed from 8.3 onwards

August 20, 2016, 9:40 am
$
0
0
Hi Guys,

Just want to know, Is "lun alignment show" command removed from 8.3 onwards.

regards
Vinay
↧

Why 8.3.x has betther performance than 9.0?

August 21, 2016, 8:14 pm
$
0
0

Hi all.

 

I simulated by spm.

 

Same Model FAS8040A, same workload - 80% random read 20% random write, same protocol and same disks.

 

But Ontap 9.0 has poor performance Maximum IOPS and throughput.

 

I only changed version of Ontap!

 

Below pictures are performance results at simulated by netapp SPM.

 

Left image is Ontap9

 

11111.png

↧

vsadmin/user authentication against AD via ldap only?

August 22, 2016, 4:29 am
$
0
0

Hello all,

 

I'm currently trying to get vsadmin authentication, against AD, working without using a CIFS enabled SVM. My LDAP queries seem to be working just fine when debugging things with secd. However I just can't get the actual logging in to work for the other engineers.

 

Is what I want even possible? All I've found on what I want if at the bottom of page 117 of the ontop sysadmin guide for 8.3. However what I'm actually doing is LDAP authentication that happens to be coming from my AD server (could just as well be openldap), do anyone have an idea what could be happening here?

 

The error I'm encountering is:

clust-dr-1::*> diag secd name-mapping show -node clust-dr-1a  -vserver clust-dr-1-backup -direction unix-win sjoerdoo

Vserver: clust-dr-1-backup (internal ID: 2)

Error: RPC map name request procedure failed
  [  1 ms] Using a cached connection to <removed ip>
  [     3] Trying to map 'sjoerdoo' to Windows user
           'MYDOMAIN\sjoerdoo' using LDAP
  [     4] 'CifsServerSecurity' configuration not available
  [     4] Could not find Windows name 'MYDOMAIN\sjoerdoo'
**[     4] FAILURE: Name mapping for UNIX user 'sjoerdoo' failed
**         with transient errors.

Error: command failed: Failed to find mapping for the user. Reason: "SecD Error: The mapping operation
       failed due to some transient failure".

The 8.3 cdot admin guide says:

 Authentication methods for user accountsˀWindows Active Directory authentication (domain)

    • For Windows Active Directory authentication, a CIFS server must be created for the Storage Virtual Machine (SVM), and Windows domain users or groups must be mapped to access-control roles by using the security login create command with the -authmethod parameter set to domain for the cluster and SVM access.
      In addition, to authenticate Windows Active Directory domain users or groups for cluster access, a tunnel must be set up through a CIFS-enabled SVM.

↧

Netapp ONTAP 8.3.1. NFS hardening

August 22, 2016, 5:37 am
$
0
0

Hi,

 

we have a nice netapp cluster with 8.3.1 running.

We have multiple vservers for NFS iscsci and CIFS. I am running into the following problem.

A linux coworker of mine is able to mount all the NFS volumes on my filers within /

We have NFS export policies enabled with allows servers in 2 vlans with acces to certain mounts.

However, my coworker can mount / and see all the mounts on the filers.(because he is in one of the 2 vlans)

How can I disable this? The volumes are all mounted under namespaces under /.

So if I remove the export rights of / all the other volumes beneath / will also be unmountable?

 

thanks!

↧
↧

is Nc-RenameExportPolicy Disruptive operation ?? - Cdot

August 22, 2016, 8:28 am
$
0
0

Hi Team,

 

Need a small help.

We've got a requirement to change export-policy names to all of our C-Dot filers.

After Observing, We came to know that they are 1000's of export policies needs amendment

 

We've had a powershell script in place to change the export-policy as per our requirment

 

Now just we need a small confirmation, Does Changing export-policy name(not underlying its rule index) will create any impact to the on-going process ?

 

Please let me know as soon as possible this change is scheduled in 2 days.. and really dnt want to take any chance in fail. 

↧

Windows report incorrect usage of CIFS share

August 23, 2016, 1:41 am
$
0
0

Hello Colleagues,

Im facing issue, that confusing me, and even more our customer. Output of "df -g" and "df -sg" command show, that there is 14TB on volume and qtree shared over CIFS. Same show also "quota report".

Point is, that customer on host, where drive is mapped, once open properitied, there is only 2,1TB utilization shown.

Any idea, what is the problem? I beleive its problem of windows, anyway someone of you might give me some help here hopefully.

 

Petr Jaros

↧
Search

Problem assigning replacement disk

August 23, 2016, 2:14 am
$
0
0

We had a faulty disk and recieved a replacement from NetApp.

But when I swapped out the faulty disk and inserted the replacement, the amber warning light conitinued to display.

 

When I run storage show disk -p

 

I see all of my disks and their partner but the faulty disk name's partner field is blank.

 

A message says, "1 disks have downrev firmware that you need to update.

 

When I tried disk assign all

 

The error says, "Assign faield because disk 'diskname' has redundancy errors and force option is not specified"

 

We haven't tried the force option yet.

 

Can anyone please advise what could be causing this & how we can resolve it? Thanks!

 

 

↧

SVM DR on 8.3.2

August 23, 2016, 3:42 am
$
0
0

Hi, I'm looking to setup SVM DR on a four node cluster. The idea would be to have the production SVM hosted on two of the nodes, and the DR SVM hosted on the other two nodes., all within the same cluster. Does anyone know whether this will work or does this feature only work between two separate clusters? 

 

Thanks!

↧
↧

Cifs name lookup failed with dynamic home drives

August 23, 2016, 8:04 am
$
0
0

Ok, so i'm confused... I do the following

 

show-user -vserver vserver1 -username myusername

 

Vserver: vserver1  (internal ID: 3)

Error: Lookup of CIFS account name procedure failed
[ 1 ms] Successfully connected to ip 10.1.1.1, port 445 using
TCP
[ 9] Successfully authenticated with DC
mserver.mydomain.com
[ 12] Could not find Windows name 'localvserver\myusername'
**[ 12] FAILURE: CIFS name lookup failed


Vserver : vserver1
Username : 'localvserver\myusername'

 

Why on earth is it pointing to my local vserver and not my domain.

 

It should come up as 

 

mydomain\myusername

 

I'm confused as to why this is happening, and I missing an option.  secd returns authentication fine, and my HD mounts fine.  I'm running 9.0RC2,  I believe this started in 9.0

 

 

↧

SNMP volume instance

August 23, 2016, 2:54 pm
$
0
0

I'm using WhatsUp Gold to monitor the sizing of the volumes on my filer running ONTAP 8.2.4 7-Mode. All the volumes have the same SNMP OID and are only distinguishable by their instance number. The problem is that every time the filer gets rebooted, the instance numbers on the filer get scrambled and I have to remap the SNMP monitor with the new instance numbers. Is there a way to make the volume instance numbers static, even after a reboot?

↧

"operation not permitted" error when copying multiple files

August 25, 2016, 1:46 pm
$
0
0

FAS3240 running 8.1.4P1 7-Mode exporting NFS shares to Oracle Linux 6.7 (just like Red Hat).  NetApp is connected using Fiber Channel and nfs 3.

When copying multiple files from script we will occasionally get from the Linux servers:

 

cp: cannot create regular file `/directory/subdirectory/filename':Operation not permitted

 

This will not do it for all the files, just intermittently on files as the script runs.  And usually if we get the error above the file will be created but will have 0 byte count. 

We haven't seen it happen if we run copy commands individually from command line.  It appears to happen when copying multiple files using a script.  We're talking about 890 files.  Most are not large.  2 of them are 12MB in size.  Most of the others range from 10 to 20K.

 

Nothing in the messages file or in any of the log files that I've found.

 

I'm wondering if there is something going on where the filer can't  keep up with the multiple file copies.

 

Any ideas where to start looking to troubleshoot this?

↧
Viewing all 4994 articles
Browse latest View live
Search
<script src="https://jsc.adskeeper.com/r/s/rssing.com.1596347.js" async> </script>