Thursday, October 8, 2015

Netapp Load Share Mirroring For Volume

What are Load Sharing Mirrors

One can create a load sharing mirroring relationship on a FlexVolume which has high read hits. We can have one or more load sharing volumes for a source volume where all the read's would be directed to the destination read-only volumes which acts as load sharing for the source volume.

To create a load Sahring volume we have to choose -type parameter as DP, Now i have volume Manufacturing where i will be creating two load sharing volumes (Mirror1 & 2)

Nayab_C01::> volume create -vserver vs0 -volume Manufacturing_mirror1
-aggregate aggr2 -size 3GB -type DP

Creating one more load sharing mirror volume

Nayab_C01::> volume create -vserver vs0 -volume Manufacturing_mirror2
-aggregate aggr2 -size 3GB -type DP

After we have created Load Sharing mirror volume we have to establish relationship with the designated volume to direct all READ'S to the load sharing volumes and when configuring Snapmirror we have to choose -type as LS


Nayab_C01::> snapmirror create -source-path //vs0/Manufacturing
-destination-path //vs0/Manufacturing_mirror1 -type LS

[Job 171] Job is queued: snapmirror create the relationship with destination
[Job 171] Job succeeded: SnapMirror: done


Doing same for mirror2

Nayab_C01::> snapmirror create -source-path //vs0/Manufacturing
-destination-path //vs0/Manufacturing_mirror2 -type LS

[Job 201] Job is queued: snapmirror create the relationship with destination
[Job 201] Job succeeded: SnapMirror: done

Once we had the relationship established now we have initialize the baseline copy for Load Sharing

Nayab_C01::> snapmirror initialize-ls-set //vs0/Manufacturing

[Job 211] Job is queued: snapmirror load-share initialize for source //vs0/Manufacturing


Now we will have all our READ's redireced to the loadsharing destination volumes

Things to Remember :-

1) A load sharing volume should be mounted under the same SVM namespace as the source volume

2) Load Sharing volume is not supported for Infinite Volume

3) Load Sharing mirrors should be created on different aggregates and accessed by different node in the cluster to achieve load balancing

4) NFSv4 clients are not supported with read-only load-sharing mirrors, Inorder all the traffic from      NFSV4 clients will be routed to source volume from READ's and WRITE's




Monday, September 21, 2015

Netapp Cluster Mode Maximum Volume parameters





Parameter
Maximum Value
Maximum Volumes/Node
500
Maximum Volumes/24-Node
1,200
Minimum Flexible Volume
20MB
Maximum Flexible Volume ( 32-Bit )
16TB
Maximum flexible volume ( 64-Bit )
50TB to 100TB

Wednesday, September 9, 2015

How to search or view Events, Alerts or notifications in Netapp Cluster Mode

event log show -event *<keyword>*

For Ex:- event log -event *disk* ( This will show all the events with respect to disk )

event log show -event *<phrase>*

Ex:- event log show -event *"on-disk mailboxes"*

event log show -messagename *<msgname>*

Ex:- even log show --messagename kern.uptime.filer ( This will all the events with respect to uptime of the filer )

event log show -severity *<keyword>*

Levels of Severity :- Emergency, Alert, Critical, Error, Warning, Notice, Informational, Debug

Ex:- event log show -severity INFORMATIONAL ( This will show all the informational events )

event log show -time >= timeframe 

Ex:- event log show -time >=30mins  ( This will show events for last 30mins )

event log show -node  *<nodename>*

Ex:- event log show -node cluster01-02 ( This will show events of that particular node )

Tuesday, July 28, 2015

Replacing a Mother Board in a NETAPP cluster-mode controller

Recently i have replaced a mother board in a cluster mode system for one of my customer and after replacing i have booted the controller and observed it booted in to 7-Mode then immediately realized that i need to change the BOOTARG setting

bootarg.init.boot_clustered true 

Cluster mode needs a specific programmable ROM (PROM) environment variable setting to be set as above. Once after i have changed the settings my controller booted in to cluster mode :)

Tuesday, June 16, 2015

What are interface groups in Netapp

An interface group is a mechanism to group together multiple network interfaces (links) into one logical interface (aggregate). After an interface group is created, it is indistinguishable from a physical network interface.
The following figure shows four separate network interfaces, e3a, e3b, e3c, and e3d, before they are grouped into an interface group
                         


The diagram below shows four interfaces trunked together as Trunk1 



                             

Different vendors termed it with their own unique name virtual Aggregation or Link Aggregation or Trunks or Ether Channel


Interface groups provide several advantages over individual network interfaces:
  • Higher throughput
    Multiple interfaces work as one interface.
  • Fault tolerance
    If one interface in an interface group goes down, your storage system stays connected to the network by using the other interfaces.
  • No single point of failure
    If the physical interfaces in an interface group are connected to multiple switches and a switch goes down, your storage system stays connected to the network through the other switches.

Tuesday, June 2, 2015

Simple Notes to understand Vfiler's in Netapp

Vfiler also called as virtual filer is a logical partition of network and storage resources in Ontap
To avail vfiler functionality need to install MULTISTORE license.
 
Protocols supported :- NFS, CIFS, iSCSI, HTTP, NDMP, FTP, SSH and SFTP
 
Maximum vfiler's :- 64 vfiler's can be created on a storage controller
 
Vfiler configuration are saved in volumes/qtrees where they are being created
 
Best practice is to have a volume as resource of a vfiler
 
Destroying a vfiler won't destroy data instead the volume will be moved to the vfiler0 ( Which is the default vfiler created once license installed ) . Also take note FCP protocol is supported only in vfiler0
 
Can use DATAMOTION to migrate vfiler
 
All the routing information  for the vfiler's can be viewed from the /etc/rc of vfiler0
 
One can add/delete/move resources like volumes/qtrees and interfaces between vfiler's

 

How to configure Netbackup KMS

 

Configuration of NetBackup KMS
 
 NetBackup security and encryption provide protection for all parts of NetBackup
operations. The parts that are made secure include the NetBackup master server,
media server, and attached clients. Also made secure are the operating systems
on which the servers and clients are running. The backup data is protected through
encryption processes and vaulting. NetBackup data that is sent over the wire is
protected by dedicated and secure method.
 
Here we are going to configure the tape based encryption, Please note that there is Media sever level encryption as well.

1)    Tape Library Configuration
 

a.      Application Managed Encryption (AME) does not require a key.

Library Managed Encryption (LME) and System Managed Encryption (SME) require a license key which is available by purchasing Feature Code 5900. Minimum

 

·        Prerequisites for Application Managed Encryption (AME) - LTO 6

·        SAS and Fibre Channel LTO Ultrium 6 Tape Drive (Full High or Half High)

·        Ultrium 6 Tape Cartridge

·        Library firmware level B.50 or higher

·        Drive firmware level C800 or higher, for reference go to http://www.ibm.com/fixcentral.

·        Encryption Key Manager application

b.      Tape Library:

                                                    i.     Click Configure Library > Encryption in the left navigation panel.

                                                   ii.     On the Encryption screen, select an Encryption method for each logical library. Without an encryption license key, select None or Application Managed Encryption.

                                                  iii.     Click Submit to apply the changes.

 
2)    Overriding the Symantec Intrusion Security policy (IPS)

 
a.      Use SSH to login to the appliance as an Administrator

b.      Go to Support > Maintenance menu. If you try to access the elevate command the following message is displayed:
 
Permission Denied !! Access to the root account requires overriding the Symantec Intrusion Security Policy. Please refer to the appliance security guide for overriding

c.      Run the Support > Maintenance command.

d.      To enter your Maintenance account, run the following command, and provide the password when you receive a prompt.

 NBAppl.Support > Maintenance

<!--Maintenance Mode--!>

maintenance’s password:
 
e.      In the Maintenance mode, type the following command to override the Symantec

Intrusion Security Policy:

/opt/Symantec/scspagent/IPS/sisipsoverride.sh

 To override the policy and disable protection, enter your login password.

Password:

f.       Enter your maintenance password.

 The appliance then displays the following options:

Choose the type of override that you wish to perform:

1. Override Prevention except for Self Protection

2. Override Prevention Completely

Choice?

 g. Enter 1 to override prevention except for self protection.

 The appliance displays the following options:

Choose the amount of time after which to automatically re-enable:

1. 15 minutes

2. 30 minutes

3. 1 hour

4. 2 hours

5. 4 hours

6. 8 hours

7. never

 h.      Enter the appropriate number from 1 to 7 based on the time required to debug the Symantec support case.

The appliance displays the following message:

Enter a comment. Press Enter to continue.

i. Enter a relevant comment as to why the override is required.

 The appliance overrides the policy and displays the following message:

Please wait while the policy is being overridden.

........

The policy was successfully overridden.

 maintenance - !> elevate

 You should now have access to the root account for debugging the appliance.

 3)    Installation and Configuration of KMS
 
a. Run the following command to change directory:

 maintenance - !> cd /opt/openv/netbackup/bin

 b. Run the nbkms -createemptydb command.

c. Enter a passphrase for the host master key (HMK). You can also press Enter to create a randomly generated key.

d. Enter an ID for the HMK. This ID can be anything descriptive that you want to use to identify the HMK.

e. Enter a passphrase for the key protection key (KPK).

f. Enter an ID for the KPK. The ID can be anything descriptive that you want to use to identify the KPK.

g.The KMS service starts when after you enter the ID and press Enter.

h. Start the service by running the following command:

 nbkms

i. Use the grep command to ensure that the service has started, as follows:

 ps -ef | grepnbkms

 j.       Create the key group. The key group name must be an identical match to the volume pool name. All key group names must have a prefix ENCR_.

 To create a key group use the following command syntax.

 nbkmsutil -createkg -kgname ENCR_volumepoolname

 The ENCR_ prefix is essential. When BPTM receives a volume pool request that includes the ENCR_ prefix, it provides that volume pool name to KMS. KMS identifies it as an exact match of the volume pool and then picks the active key record for backups out of that group.

k.      Create a key record by using the -createkey option.

nbkmsutil -createkey -kgname ENCR_volumepool -keyname keyname -activate -desc "message"

 The key name and message are optional; they can help you identify this key when you display the key. The -activate option skips the prelive state and creates this key as active.

l.       Provide the passphrase again when the script prompts you.

 In the following example the key group is called ENCR_pool1 and the key name is Q1_2008_key. The description explains that this key is for the months January, February, and March.

 nbkmsutil -createkey -kgname ENCR_pool1 -keyname Q1_2008_key -activate -desc "key for Jan, Feb, & Mar"

 m.    You can create another key record using the same command; a different key name and description help you distinguish they key records:

 nbkmsutil -createkey -kgname ENCR_pool1 -keyname Q2_2008_key -activate –desc "key forApr, May, & Jun"

 Note: If you create more than one key record by using the command nbkmsutil -kgname name -activate, only the last key remains active.

 n.      To list all of the keys that belong to a key group name, use the following command:

 nbkmsutil -listkeys -kgname keyname

Note: Symantec recommends that you keep a record of the output of the nbkmsutil -listkeys command. The key tag that is listed in the output is necessary if you need to recover keys.

 

 

Monday, March 30, 2015

Power Off procedure for BROCADE Switches and Directors

 


To avoid corrupting your file system, it is recommended that you perform graceful shutdowns of switches and Directors.  

For Directors running Fabric OS versions prior to 5.1.x, the following procedure describe how to gracefullyshut down a Director:

1.Verify which CP is the active CP, and log in to the active CP using a Serial Console connection.

2.On the standby CP, set the slider switch to the off position, or eject the standby CP from the chassis. Thisdisables the standby CP.

3.Enter the reboot command from the active CP. This will gracefully take down the system.When you see the "Press escape within 4 seconds to enter boot interface" message, press ESC to

4.suspend the active CP.

5.Power off the chassis by flipping both AC power switches to "0" (LEDs inside AC power switches shouldturn off). To maintain the ground connection, leave both power cords connected to the chassis and to electrical outlet.


For both switches and Directors running Fabric OS 5.1.0 and later, it is recommended that you use the following graceful shutdown procedures


1. Connect to the switch and log in as admin.

2.Enter the sysShutdown command.At the prompt, type y.

3.switch:admin> sysshutdownThis command will shutdown the operating systems on your switch.You are required to power-cycle the switch in order to restore operation.Are you sure you want to shutdown the switch [y/n]?y

4.Wait until the following message displays:Broadcast message from root (ttyS0) Wed Jan 25 16:12:09 2006...The system is going down for system halt NOW !!INIT: Switching to run level: 0INIT: Sending processes the TERM signal Unmounting all file systems. The system is halted flushing ide devices: hda Power down.

5. Power off the switch

Monday, February 16, 2015

Prerequisites for GRT Backup/Restore NetBackup 7.5 and Windows 2008 R2 Domain Controller

Pre-Requisites

NFS must be installed on Media Server and All AD Domain Controllers which are to be backed up with GRT enabled

Install NFS Services on Media Server and All the Active Domain Controllers Windows 2008/R2

For NFS, Add File Service Server Role

Select Services for Network File System as Role Service , Complete the wizard

From Media Server configure portmap service to start automatically after server restart

From CMD execute “sc config portmap start= auto





From Service.msc console Stop and Disable below services

Server for NFS on both Active Domain Controller and Media Server And Disable Client for NFS on Media Server

If Active Domain controller is also Media Server disable both Client/Server for NFS Service
Create No.Restrictions Touch (empty) File on Media Server under /veritas/netbackup/db/altnames

If altnames directory does not exist, create it

Make sure Touch file does not have suffix of .txt it should be No.Restrictions only

To rename extension of file, from windows explorer navigate to Folder Option (Alt+T+O)

Folder options=> View => Uncheck Hide Extensions for known file types => Apply => Ok

Make sure you revoke above change after file is created.

Privileged Account for GRT

NetBackup Client Services at Active Domain Controller must be running with Domain Admin Privileged Account



Tuesday, February 10, 2015

How to identify the disk name in Netapp

We all pretty well know every disk will have a universal unique identifier UUID or serial number , but  each disk will have a unique name depending on how they are connected to the storage system.

Note :- Always remember For internal disks, the slot number is zero, and the internal port number depends on the system model.

First knowing the basics each disk shelf will have bay's to hold the disk

Ex:- A Netapp DS4243 is a 24 bay disk shelf which is 4U in size.

Now each disk shelf will be identified with a name based on the Shelf ,Bay and the port to which the shelf is connected and we have different methodology for different type of DISK CONNECTION
                 
For the SAS, Direct Attached type of disk connection the disk name will  be based on

<slot><port>.<shelf ID>.<bay>  ( In short i remember it as SPSB ) 


Ex:- Which means say if i have a disk in bay 12 of Shelf 1 which is connected to onboard port A

My disk ID would be :- 0a.1.12 

For the FC-AL, direct-attached type of disk connection the name is based on

<slot><port>.<loopID> ( In short i remember it as SPL )

Ex:- If my disk loop ID is 20 connected to expansion card on slot 8 to port C

My disk ID would be :- 8c.20

For FC-AL , Switch attached type of connection the name is based as below

<switch_name>.<switch_port>.<loopID>


Ex:- If my disk loop ID is 40 connected to port 4 of a switch 4

My Disk ID would be :- SW4.4.40








  

Disk Physical Size Vs Usable Size


Wednesday, January 7, 2015

Tool to analyze Netapp Perfstat Output

Wondering how to analyze Netapp Perfstat , The tool below will help you out to know the CPU , Disk Overhead etc...


1) Download and run the .exe from the link above 
2) It will open a your command prompt in windows machine to key in the Perfstat file name 

3) Enter the perfstat file name with complete location





 4) Now enter the controller name for which you want to analyze the perfstat





















This will generate three text files for NETWORK , CPU & DISK in the same folder where your perfstat_analyzer.exe was present 

Friday, January 2, 2015

Difference Between Netbackup and Backupexec

Well we know Netbackup is for Enterprise level and Backup is for Mid-sized environments , I would like to highlight few more difference's which are mainly considered while discussing about these two leading backup products from Symantec.
  
Category
Backup_Exec
Netbackup
Latest Version
2014
7.6.0.2
Database in the backend
Microsoft SQL
Sybase
Tape Format
BKF
TAR
Reporting

Native
OpsCenter
Management
CASO (Central Admin Server Option) can manage multiple media server deployed across Domain.
NetBackup Master Server, can manage multiple media server, SAN media servers and clients centrally.
Can read from NetBackup/Backup Exec

Not supported to read NetBackup images to Backup Exec
Supported until BE 2012 , later versions not supported
NDMP

Supported
Supported
De-Duplication
Supported
Supported
GRT
Supported
Supported
BMR
Supported
Supported
Tape Library, VTL and autoloader
Supported
Supported
Tape-out protocols FC & ISCSi
Supported
Supported
Multistreaming backups

Not Supported
Supported
Multiplexed backups

Not Supported
Supported
Vmware and Hyper-V Environments
Supported
Supported
Offline and Online backups
Supported
Supported
Automatic scheduled and Manual Backups
Supported
Supported
Media Server Support
Windows Only
Windows as well as UNIX
Client Support
Windows and Linux Only
Windows as well as UNIX
Disk Based Backups
Supported
Supported
             Encryption

Supported
Supported
SAN backups
Supported
Supported
Convert backups to virtual machines
Supported
Supported