Tivoli Storage Manager 7.1.1 Technical details

Report
Tivoli Storage Manager 7.1.1
Server, Client, Operation Center, HSM & FB4W
Tricia Jiang
IBM Tivoli Software |
© 2014 IBM Corporation
TSM 7.1.1 –New Features
Operations Center
 Daily Email Reports – executive summary
 Alert Configuration
 Client, Admin and Server Maintenance schedule
 Storage Pools details
 Bookmarks
 Services (Policies) Details
 SUR Occupancy
 Client Management Services
TSM Server
 Replication: Scale Enhancements
 Replication: Dissimilar Policies
 Replication: Repair/recover files for source server from
replication target
 Compress DB2 backups & archive logs
TSM Client
 256-bit AES encryption
 Conformance to NIST SP800-131a
 Windows 2012 R2 support
 Solaris x86 BA client (64-bit)
HSM for Windows
 Backup Performance Enhancement
TSM for Space Management
 GPFS User Exit Based Failover
 UNIX: Enhanced Recall
Fastback for Workstations
 Central admin console Websphere Liberty Profile
 Graphing in the reporting tables in Central admin console
 New block level backup strategy
 Help system upgrade
2
GA 9/12/14
TSM for VE
 Provide initial TSM for VE solution for Microsoft Hyper-V
 Integrate simplified mount solution for file-level recovery in DP for
VMware UI; "Help Desk"
 FCM & DP for VMware - integrated vSphere Web Client user interface
(Adobe Flex)
TSM for ERP 6.4.2 (shipped July 25th)
 Performance Enchantements
 DB2 10.1 /10.5 support
 Support BRTools 7.4
 Windows 2012 Support
 Solaris 11 support
TDP (Currency)
 Windows 2012 R2 - Exchange, SQL, Oracle,
 SQL 2014 support
TDP Exchange
 Windows 2012 R2 support
 Reduced authority for Exchange IMR – Roll Based Access Control support
 IMR Usability and Performance enhancements
 Exchange config wizard to validate permissions
FCM for Windows:
 SQL Server 2014 Support, Windows Server 2012 R2 Support
 Exchange – IMR Browser Usability & Performance see WI 3304 above
FCM for VMware:
 GUI integrated with vSphere Web Client
 Configuration Wizard
 NIST SP 800-131a compliance
 Application consistent FlashCopies from target of a metro/global mirror
TSM Operation Center - Dashboard or Overview Changes
What’s new in 7.11
 Multiple videos on what is new
 Shows each time OC started
 Can Click “Don’t show again”

Polices graphics
Expiration
The 7.1.1 Operations
Center dashboard
areas of change :
3
Traffic Graphics
Customizable Traffic view
Usage of
DB & Log
Space
TSM Operations Center – Daily Email Report
•
•
•
4
Executive summary of over health and client coverage
Daily Email Reports
• Summary: Backup Success,
Server & Storage availability,
Client Protection
• HTML detailed report:
Summary, At Risk Clients, TSM
Servers, Storage Pools, Storage
Devices
Works with most emails & browsers
Email Configurable
• Recipient (TSM admin or email)
• Time of day sent
TSM Operations Center - Daily Email Report
• Daily email that summarize the basic health of your TSM environment:
• High level summary: Backup success, Server and storage availability, Client protection
• Attached detailed report: Summary & Details: At Risk Clients, TSM Servers, Storage Pools, Storage Devices
• Requirements:
• Setup a SMTP server that allows open mail relay
• No authentication, No SSL, OC must have access to
• SMTP server address and port
• TSM administrator with email address associated
• update admin {admin_name} [email protected]}
• Email message:
• Sent in HTML format
• Viewable in email clients that support HTML format
• Lotus Notes, Outlook, Apple Mail, Gmail and others
• Might not be shown correctly in email clients that convert HTML to another format
• For example, if HTML email is converted to Rich Text Format (RTF)
• Attached detailed HTML report
• Download report and then open the HTML file with web browser
• Email client preview feature not supported
• Report will open but because of security features report may not display or function correctly
5
TSM Operations Center – Client Schedules
•
•
•
•
6
Visualize and Create Client Schedules
View all client schedules
across servers
Create New Schedules
Edit existing schedules
See client associations
TSM Operations Center – Create New Client Schedules
7
Visualize and Create Client Schedules
Wizard
• Different prompts for schedule
type and subtype
• Followed by client association
TSM Operations Center – Client Schedules Details
•
•
•
•
8
Visualize and Create Client Schedules
View schedule start window
Visualize start window, duration
Modify schedule attributes
View and mange client
associations to schedules
TSM Operations Center – Maintenance Schedule
•
•
•
View Maintenance activities for servers
• Read only
• 2 week history
“Initiated by” = how task started
• Automatic Tasks
• Commands
• TSM Schedule
• External
New “end time” for admin schedules
•
•
9
Visualize and Create TSM Maintenance Schedules
MAXRUNTIME parameter in DEF SCHE
Server processes still running past this
time are cancelled by the server
Maximum Run Time on Client & Administrative Schedules
Description
• For client schedules
• TSM Admin can specify the max amount of time a schedule runs before getting a warning message
• For administrative schedules
• TSM Admin can specify the maximum amount of time the schedule can run
• If the schedule has reached the maximum time, it will get cancelled
Benefit
• Administrator can better control the scheduling of client and administrative tasks
Server Support
• MAXRUNTIME is a new parameter added to DEFINE SCHEDULE command
• Default is 0: no maximum run time
10
TSM Operations Center - Storage Pool
Storage Pool Configuration
• Capacity, Next Pool, Start / Stop Migration, Start Reclamation, # of Volumes
2 week storage pool activity History
• Activity Graph + daily Backup/Reclaim/Migrate operation completion
Edit storage pool properties
View storage pool volumes
• Capacity, Access, Online/Offline or Status, Reclaimable Space capacity
Visibility for Reclaimable Space
• All STGpool Table: Reclaimable Volumes added as hidden table
• Volumes Level: % Reclaimable Space & Reclaimable Space Capacity
11
TSM Operations Center - Services Policy Details
•
•
Last Activated
Will display
Migration
Destination
here if set
View when policy domain was
activated in the last 2 weeks
7.1.1+ servers ONLY
Visual information on Storage Pools
2 week expiration history
Policy Attributes
• Retention Settings
• Versions settings
• Pool destination
• Activation
TSM Operation Center - Reclaim Storage Pool Action
•
•
13
Kick off Reclamation from the OC
Ability to manually kick off
Reclamation directly from the
OC
Underneath the More button
in Storage Pools
Volumes number might change
as you slide the Start
Reclamation threshold
TSM Operation Center - Migrate Storage Pool Action
•
•
14
Kick off Migration from the OC
Ability to manually kick of
migration directly from OC
Underneath More button in
Storage Pools
TSM Operations Center - Advanced Filtering
Magnifying glass icon opens advanced filter
Build a custom search of one or more search terms

Column Name - Condition - Content

Use “+” “-” buttons to add and remove terms

Use “And” “Or” buttons to switch between joining operators
 When an advanced filter is applied
• Reset button appears
• Basic filter displays “Advanced Filter ON”
• Column headers being searched are shown as depressed
15
Customize items being viewed
TSM Operation Center - Global Shared Links
Description: List of shared links that can be used by all OC users
Benefits: Share commonly used links or sorts
OC Specifics:
 Admin Authority
– System Admin: add, delete, update, view & open links
– Non-system admin: viewing and opening links
 To control entries, click enabled save button
– Add
• 10 entries in maximum can be added
– Delete
• Any existing entries can be deleted
– Update
• Both Name and URL entries can be updated
– Opening the links
• OC internal links will be opened in the current tab
• Others will be opened in a new tab
 Dropdown list in banner
– Open the existing shared links
– OC internal links will be opened in the current tab; others will be
opened in a new tab
– Launch the shared links tab of global settings dialog
16
TSM Operations Center - Client Management Services
View and understand client
problems directly from OC
• SW runs on the client machine
• Retrieves errors from client’s
DSMERROR.log
DSMSCHED.log
17
Visual information on Storage Pools
TSM Operations Center - Client Management Services
• Benefit
• Enables TSM admin to view & understand client problems directly in OC
• No need to log into the client machine and manually viewing log files
• Client Specifics
•
•
•
•
•
•
Client Management Service (CMS) is software that runs on the client machine itself
Supports TSM Backup-archive client system at version 6 or 7 (not tested on TDPs)
Supports TSM client OS: Windows 32 & 64 bit & Linux 64 bit
In OC, TSM admin needs to have: System, Policy, Client Owner or Client Access authority
Need to make sure firewalls are configured for all initiated connections
Can process up to about 500 MB combined dsmsched.log and dsmerror.log
 GUI Install Steps (approx 5 minutes)
• Obtain installation package from product DVD or IBM download site
• 7.1.1.000-TIV-TSMCMS-Linuxx64.bin
• 7.1.1.000-TIV-TSMCMS-Windows32.exe
• 7.1.1.000-TIV-TSMCMS-Windows64.exe
• Run the install package in a directory; this will extract the installation files.
• Run the install.bat (Windows) or install.sh (Linux) to initiate installation

Silent Install Steps (approx 2 minutes after file created & edited)
•
•
•
•
18
Copy the install_response_sample.xml to install_response.xml
To change port install_response.xml file line: <variable name='port' value='9028' />
On Linux./install.sh -s input input/install_response.xml -acceptLicense
On Windows install.bat -s input input\install_response.xml -acceptLicense
TSM Operation Center - Count of Reclaimable Volumes
Description
• Estimated count of reclaimable volumes can be retrieved from the server by the Operation Center.
Benefit
• OC’s storage pool view shows the number of reclaimable volumes for each sequential type storage pool
Server Support
• The count includes volumes that meet the following criteria
• Volume has a percent reclaimable value that is equal to or greater than the pool's reclaim threshold
• Volume access is READWRITE, READONLY, or OFFSITE
• For spoke servers that are unavailable at the time of status monitor collection
• The number of reclaimable volumes will not be available and will show zero value for that server
• For spoke servers that are downlevel (prior to 711)
• The NUM_RCLMVOLS value is blank (null)
19
TSM Operation Center – Customizable Logon Screen
• Customize the TSM OC logon to highlight information important to your users

Custom Login Text
 Create or Update <install dir>/ui/Liberty/usr/servers/guiServer/loginText.html
 Can add custom login text that will be shown on the login screen
 Can add html tags
20
Normal –vs- Customized
TSM Operation Center - Automatic Refresh of the GUI Data
Description
• Prior to V7.1.1, the server refreshed data in the GUI table at the status refresh interval (default 5 mins)
• Starting with V7.1.1, the server will immediately refresh the data on server tables used by the OC
Benefit
• Allows the OC to display more up to date information without waiting for status refresh interval
Server Support
• Applied to all status monitor table( TSMGUI_nnnn_GRID tables )
• Currently does not apply to Alerts table – it has it’s own update and refresh handling
• Does not apply to TSMGUI_STATUS_DATA
• Table used by OC “Overview” page
• Refresh is still done only at status refresh interval due to constant changing nature of the data
• Full refresh of data in all OC related tables ( TSMGUI_xxxx_xxxx) are done at status refresh
• No auto refresh if status refresh interval is set to 3 minutes or less
• Avoids conflict between the two refresh cycles
• There’s minimal delay, for example if a node is added to the schedule in the OC
• Update is made on the server; Server pushes the update to the OC; OC table gets refreshed
• Status Retention is changed from 48 hours to 24 hours
21
TSM Operation Center - Alert Updates
 Alerts were introduced in the 6.3.4 TSM server
 7.1.1 Operations Center
–
–
–
–
–
Define new alert trigger on Hub server and on any Spoke servers for which alert trigger is valid
Change category of alert triggers on Hub server and on any Spoke servers on which alert triggers exist
Remove alert triggers from Hub server and from any Spoke servers on which alert triggers exist
Configure alert email notifications for administrators
Does not configure hub server and spoke servers for email notification
Click Configure to
change alert settings for
the Hub and any Spoke
servers
Can change the
category the alert
belongs to
22
TSM Server and OC SUR Occupancy
Description
• TSM server collects the space occupancy of primary storage pools for the purposes of SUR licensing
• Keeps a record of the space occupied by primary pool storage pools supports TSM SUR
• Suite for Unified Recovery is licensing based on the amount of data being managed
Benefit
• Administrators don’t have to run a long SQL statement to get the SUR information
Server Support
•
•
•
•
•
Collected once a week
Logged to the Activity Summary table as ACTIVITY=SUR_OCCUPANCY – in TB or PB
QUERY STATUS is updated to display the value and time when it was last collected
SELECT * from STATUS will also display the value and time
Can view in OC’s Server view by adding in the SUR Occupancy and SUR Collected columns
QUERY STATUS output
tsm: FUSION>q status
Storage Management Server for AIX - Version 7, Release 1, Level 1.0
Server Name: FUSION
…
…
SUR Occupancy (TB): 4.56
SUR Occupancy Date/Time: 08/10/2014 14:21:32
SELECT output
tsm: FUSION>select TOTALSUROCC_TB,SUROCC_DATE from status
TOTALSUROCC_TB
----------------------4.56000000000000E+000
23
tsm: FUSION>
SUROCC_DATE
--------------------------2014-08-10 14:21:32.000000
TSM Server Scratch Pad
Description
• New commands allows a user to enter data on a new line in the scratch pad
• Scratch pad is a database table that the TSM server hosts
• Can use the scratch pad to store diverse information in table format
Benefit
• Allows the OC (or other TSM components like clients, or even end users) to store a wide range of data
• Currently OC use of scratchpad is not exposed
• Using internally to save some settings, filter information and link info
• Admins can use it to save some notes to themselves or for their team
Server Support
• New commands introduced
DEFINE SCRATCHPADENTRY - create a new line entry in the scratch pad
DELETE SCRATCHPADENTRY – delete an entry from the scratch pad
QUERY SCRATCHPADENTRY – query entries in the scratch pad
UPDATE SCRATCHPADENTRY – update an entry in the scratch pad
SET SCRATCHPADRETENTION – set how long to keep an entry in the scratch pad
• Modified command
QUERY STATUS – new output field added to show the scratch pad retention setting
24
TSM OC - Visualization of Down Level & Unavailable Servers
Unavailable server in table of servers
• Server status is shown with unavailable icon
• Last known data for server is presented, e.g. Used and
Free Space for Database
Unavailable server in other tables
• Last known data for server is presented, e.g. Capacity
and Access
• The row of an unavailable server is grayed out & no
actions can be performed
7.1.1 adds checks for beta code
• Will remember code levels used with a particular database
• If any level was pre-release, then ANR4901E issued at startup and on 24-hour loop:
ANR4901E The server database has been used with a pre-release server_version Tivoli Storage Manager server (driver
driver_number, build build_id). The server is not supported for commercial or production use.

Only able to detect pre-release use after 7.1.1
New SHOW VERSIONHISTORY command implemented to display version history
>show
versionhistory
Date (GMT)
-------------------2013/11/09 23:46:08
2013/11/09 23:57:31
2013/11/10 00:15:56
25
Version
-------------------7.1.1.0
7.2.0.0
7.2.0.0
Pre-release Driver
-------------------0300
• Included in output for QUERY SYSTEM
• Can be issued by any administrator regardless of authority
• If non-NULL value in 'Pre-release Driver' column, the server instance is not supported for production use
TSM Operation Center - Updates to Logging/Tracing
User can control logging from a panel in the Operations Center
Logs are centralized in same file as Admin API
26
TSM Operation Center - Updates to Progress Dialog
Task is in-progress
• Progress bar blue
• Can Hide and Show details (Details are hidden by
default)
• Check mark icon next to successful steps
• Spinner icon next to in-flight step
Task is still in-progress, but a step has
failed
• Progress bar blue
• Error icon next to failed steps
Successful completion, summarizes the
results of the task
• Progress bar green
• Succeeded text and icon
• Check mark icon next to successful steps
27
TSM Operation Center - Additional Changes in 7.1.1
• The Java Level has been upgraded
• To: Java(TM) SE Runtime Environment (pwa6470sr6fp1-20140108_01 (SR6 FP1))
• The Liberty Web server has been upgraded
• To: WebSphere Application Server 8.5.5.2 (wlp-1.0.5.cl50220140403-1858)
• The Knowledge Center upgraded to Customer Installable 1.5
• Fully supported
• Supports newer versions of FireFox, Chrome and Internet Explorer
• The 7.1.1 Operations Center requires the Hub to be a 7.1.1 TSM server
• Spoke servers are permitted to be at version 6.3.4 through version 7.1.1
• Spoke server can not be at a higher version (any part of V.R.L) than the Hub server
• Down level Spoke servers are reflected as such in the Operations Center
• If a user attempts to use less than a 7.1.1 server as the Hub for a 7.1.1 Operations Center
• Following just an upgrade of the Operations Center, the login will fail with an error indicating the Hub
server must be 7.1.1 or higher
• If the user attempts to Configure the Hub server for the 7.1.1 Operations Center, the Configuration
login will fail with an error indicating the Hub server must be 7.1.1 or higher
• Neither the TSM server install nor the Operations Center install prevent the user from upgrading one
and not the other
28
Activity Summary Record – ACTIVATE POLICYSET
Description
• New record logged to activity summary table for ACTIVATE POLICYSET command
Benefit
• Allows the OC to inform the user of when the active policy set was activated
Server Support
• Activity summary record logged when ACTIVATE POLICYSET command completes successfully
• Purpose of this record is for use by OC
• The following fields of the activity summary record are used
START_TIME/END_TIME <-- Timestamp of when the policy set was activated.
ACTIVITY
<---- for ACTIVATE POLICYSET this will be "POLICYSET ACTIVATION"
ENTITY
<---- Name of the domain
29
PERFORM LIBACTION – New ACTION Values
Description
• New action values, RESET and QUIESCE introduced to PERFORM LIBACTION
• Update all drive paths and all drives online in a specific library with one command
• New DRIVESONLY parameter allowed when ACTION=RESET
Benefit
• No need to issue multiple commands against each library’s drive & path to make them all online or offline
• In the future this may be utilized by the OC and allow the user to reset their drives and paths
Server Support
Syntax
PERform LIBAction <library name> Action=<DEFine| DELete | RESet | QUIesce DRIVEsonly = <No|Yes>
Action
Reset
update all drive paths and drives online
Quiesce update all the drives to offline; drive paths not updated only the drives themselves
DRIVESOnly
This new parm allowed only if ‘Action=RESET’ specified.
Yes vary drives online without updating any paths
No
vary drives and paths online
30
Querying DB2 Directly
Description
•
•
•
As of 6.1, TSM no longer provides ODBC interface for Admin SQL
Management Applications / Report Engines were told to use DB2 ODBC/JDBC
TSM DB2 Schema is not documented
• In particular, there's no way to distinguish between operational and reporting tables
Benefit
•
Have the ability to identify TSM DB2 Schema
Server Support
•
•
SHOW ADMINSQL identifies which DB2 views correspond to tables supported by Admin SQL interface
Publish technote to document usage and to discuss limitations
>show adminsql
Table Name
DB2 Name
Cols Staged Query
--------------- --------------- ---- ------ -----------------------------
Syntax
.-Format--=--Standard--->>-SHOW ADMINSQL---+--------------------+---><
'-Format--=--+-Standard-+-'
'-Detailed-‘
ACTLOG
ACTLOG
14 No
ADMIN_SCHEDULES ADMIN_SCHEDULES
21 Yes
QUERY SCHED F=GUI TYPE=ADMIN
ADMINS
ADMINS
20 Yes
Q ADMIN F=D
AR_COPYGROUPS
AR_COPYGROUPS
12 No
ARCHIVES
ARCHIVES
11 No
...
>show adminsql f=d
Table Name
DB2 Name
Cols Staged Query
--------------- --------------- ---- ------ ----------------------------ACTLOG
Col# Column Name
ACTLOG
14 No
Type
Length Null
---- ------------------------------ ---------- ------ ----
31
0 DATE_TIME
TIMESTAMP
1 MSGNO
INTEGER
10 N
4 N
2 SEVERITY
VARCHAR
1 Y
3 MESSAGE
VARCHAR
2001 N
4 ORIGINATOR
VARCHAR
6 Y
5 NODENAME
VARCHAR
65 Y
6 OWNERNAME
VARCHAR
65 Y
7 SCHEDNAME
VARCHAR
31 Y
New Maximum Size for Active Logs
Description
 TSM 7.1.0 increased maximum value allowed for ACTIVELOGSIZE
• Increased 4X from 131072 to 524288
• Not documented or announced
Benefit
 In 7.1.1 announcing the new limit
• Decision was too late for pubs update – 7.2 books will show the new limit
• Techdoc will be published to announce the higher maximum
• New max allows active logs to grow to 512GB
Server Support
 Larger logs can be enabled in 7.1.0 even though announcing in 7.1.1
 Things to watch out for
• Need to provision more space for logs
• Larger logs can lead to longer restart times if DB2 crash recovery is required
32
TSMDIAG - New Fields
Description
•
New fields for tsmdiag
Benefit
•
Help expedite & ease the gathering of appropriate doc to help resolve problemes
Server Support
• -servername is now required field on UNIX
• -dbalias – creates alias
• tsmdiag –dbgrowth –id admin –pa xxx –dbalias tsmdb5
• -crash – collects core files and runs platform tools
• -dbgrowth – runs dedupAudit.pl
https://publib.boulder.ibm.com/tividd/td/TSMM/SC32-9103-01/en_US/HTML/tsmdiag.html
33
TSM 7.1.1 Performance Enhancements – Node Replication’s Reconciliation
Prior to TSM 7.1.1, replication has always done a reconcile
●
–
Compares complete list of files between the source and target server
–
Used to synchronize the source and target servers / Makes target server look just like the source server
Reconcile in TSM 7.1.1- Examines entire list of files in a file space (much like pre 7.1.1)
●
–
Used during the initial replication between 7.1.1 servers
●
Once reconcile completes, change tracking processing takes over during the next replication
–
Automatically runs following a database restore on the source or target server
–
Can run manually using REPLICATE NODE FORCERECONCILE=NO|YES
• Synchronize source/targert files - used like an audit
• Resynchronize source/target files if the db is regressed (other than DSMSERV RESTORE DB)
Restartable – remembers where it left off if cancelled or after some catastrophic server event
–
Change Tracking in TSM 7.1.1 eliminates need to query target server for its list of files
●
–
New and change files are assigned a change identifier: when it’s stored & when meta data is updated
–
Replication only processes files with a change identifier – incremental replication
–
Replication picks up where the last replication left off
–
Improves performance for fs with lots of files
New Replication rule that enables replication based upon Domain
●
–
34
In addition to current replication by Server, Node and Filespace
TSM 7.1.1 - Node Replication with Dissimilar Policies
Description:
Decouple target from source server policy, allowing target server to manage replicated files via its own policy
●
Benefits:
Maintain more or fewer versions of replicated backup files between the source and target replication servers
Retain replicated archive files for more or less time on the target replication server
If fewer versions of files are kept or retained for less time on the target server, the amount of target server
storage required can be reduced
●
●
●
Server Support:
Different policy for retention time & number versions can exist on primary & target server
●
–
–
35
On source server first run VALIDATE REPLPOLICY
●
Default is to disable dissimilar policies
●
If disabled, the serves will act as they do today and honor the source server’s policy
On source server SET DISSIMILARPOLICIES server on
●
Once enabled, target server is notified during the next replication
●
VEREXIST, VERDELETED, RETEXTRA, RETONLY, RETVER, RETMIN – will impact what is kept
●
Expiration then runs independently on each server – target server runs expiration
TSM 7.1.1 - Node Replication with Dissimilar Policies
Backup data:
VEREXIST, VERDEL, RETONLY, RETEXTRA control when a backup is deleted
Archive Data: Only RETVER controls the length of time a file is retained for
Space Management Data: Retains the same behavior with or without dissimilar policies enabled
VALIDATE REPLPOLICY <server>
•
Must be ran prior to enabling dissimilar policies
•
Displays differences in policy settings for all nodes that exist on both source and target servers
•
Can be used to validate policy settings are correct, or to debug policy related issues
•
If no server is specified, defaults to current replication target server
SET DISSIMILARPOLICIES <server> <on|off>
•
•
Issued on source server to enable dissimilar policies for a specific target server
Must run VALIDATE REPLPOLICY for the given target server before this can be enabled
QUERY REPLSERVER
●
New field added to tell you if Dissimilar Policies is enabled for target replication server or not
Rebinding of Files on the Target Replication Server
36
●
In 7.1.0 <=, files bound to STANDARD on target server if the mgmtclass didn’t exist at the time of replication
•
With Dissimilar Policies, the mgmtclass that a file is bound to on the target server is now important
•
During Reconcile Replication, files will be rebound to the correct mgmtclass if it exists on the target server
•
Done at server level the first time a replication is run after upgrading to 7.1.1
•
Done at node level when REPLICATE NODE forcereconcile=yes
TSM 7.1.1 - Node Replication with Dissimilar Policies
Changes for Expiration
• Source server
•
If dissimilar policies enabled, source server holds onto a file until it has been replicated to the target
•
•
•
Files that have not yet been replicated, must be replicated before they can expire on the source
Files that have been updated ( active → inactive or inactive → expire immediately ) must be replicated (updated)
before they can expire
Bypass this requirement temporarily by updating the node or filespace replication rule to NONE
• Target Server
• All files are now deleted by expiration on the target server
•
Replication process no longer directly deletes files, instead deferred until expiration processing on the target server
•
The exceptions are group data and hsm data: These are still deleted during replication processing
• If dissimilar Policies is enabled, the targets policy take affect as data is ingested into the server
•
This is exactly the same way a client backup is handled
•
As a file comes in, if there is an extra version, it is set to be expired
• If dissimilar Policies is not enabled, the data's state is copied from the source serve
•
If it is active on the source, it will be active on the target
•
If it is inactive on the source, it will be inactive on the target
• When expiration runs, it will delete files as it always has
37
•
Files that are expired will be removed whether dissimilar policies is on or off
•
Dissimilar Policies just dictates how files are set to expired during the ingest of the data
•
ARCHIVE data that has passed its retention data is only expired on the target if dissimilar policies is enabled
TSM 7.1.1 - Node Replication with Recovery of Damaged Files
Description: Retrieve files that have been damaged on the source server from replication target server
Benefit: Assist with local recovery of files that have been damaged on the source server
•
•
Replaces damaged files with good copies that have been replicated
Includes files on volumes that have been updated as destroyed
Scenario:
1)
2)
3)
Server A replicates data to Server B
Storage pool volume or files on Server A is lost or data is otherwise inaccessible
Storage pool volume or files is restored from Server B to Server A
Server Support:



38
Must enable: SET replrecoverdamaged on
Additional step in replication process, if source server has damaged files in a file space being replicated
• Notifies the target server to resend them at the end of the replication
• When source replication ends, the target would start process to resend the damaged files
• Damaged files are deleted and replaced with new copies, retaining original insertion date
New parameter on REPLICATE NODE command: RECOVERDAMAGED=YES|NO|ONLY
• RECOVERDAMAGED=ONLY will not do replication, it will only recover damaged files
TSM 7.1.1 - Node Replication with Recovery of Damaged Files
•
New column in Replicated_Objects tracks damaged files
•
•
•
•
Valid values are currently 0 (not damaged) and 1 (damaged on source)
Worker thread periodically re-synchronizes the Damaged column
Thread wakes every 30 mins, only syncs if count of damaged changed
Source server initiates recovery, target server controls
•
At the end of a replication process, the source server will decide if a recovery is needed
•
If so, it will build up the recovery command and send it to the target for processing
•
Target starts modified replication process & corresponding replication recovery process on the source
•
The source & target will talk, with the target requesting damaged files and the source supplying them
• Grouped objects
•
Requires special consideration, object ID cannot be reused
•
Groups are repaired with the new object IDs during recovery
• Deduplication
•
39
Reuses existing replication code to only send needed chunks
TSM 7.1.1 - Node Replication with Recovery of Damaged Files
Does not seek out damage
•
•
Server must encounter damage through data movement (reclamation, client restore, move data, etc.)
Or volume must be marked as destroyed
Deduplication (Largest area of limitation)
•
•
When a base chunk becomes damaged, links to it are not automatically made invalid
Example:
• Node Bob backs up coolstuff.txt, which is split into 5 base chunks, stored on VOLA.BFS
• Node Jason backs up files, which links to all five chunks created by coolstuff.txt, stored on VOLB.BFS
• Administrator replicates node Jason, but not Bob
• Administrator marks VOLA.BFS as destroyed
• Replication recovery will not recover node Jason's damaged files
• Solution - run audit VOLB.BFS
• Audit Volume marks all damaged files on the volume, if the base chunks are damaged
• Client restore marks individual files as damaged, but not all files sharing the same chunks
• What if Node Bob was replicated?
• No additional action is necessary
• Chunks come back, links repaired throughout pool
Relinking limitations
•
•
•
One relink thread for all replication recovery requests
If server crashed/restarted, potential for many outstanding invalid links
Solution, run Audit Volume on all volumes potentially with replaced damaged chunks & run Validate Extents
Update Volume
•
•
If server crashes/restarted during an outstanding update, not all bitfiles marked damaged
Solution: Mark volume read-only, then back again to damaged
40
TSM 7.1.1 256-bit EAS Backup Archive API encryption
TSM now supports 3 algorithms of encryption
●
–
AES-128
–
DES-56
–
AES-256 (NEW)
TSM 7.1.1 introduced AES-256
●
–
Backup-Archive Client , API and DPs that use the API
–
Use ENCRYPTIONTYPE AES256 in dsm.opt or dsm.sys
●
41
Then rebackup already backed up and encrypted items
–
Data deduplication is not supported with encryption
–
“query backup” shows encryption being used
–
Client GUI will have new option
Meet National Standards
TSM 7.1.1 and FCM Conformance to NIST SP800-131a
Description:
Special Publication 800-131a
•
Information security standard of the National Institute of Standards and Technology (NIST)
•
Requires longer key lengths and stronger cryptography than other standards
•
Ability to enforce use of TLS 1.2 or higher when SSL is used
Benefit:
Stronger cryptography when using SSL
Server Support:
•
Product options to require use of TLS 1.2 or higher ( when SSL is in use )
•
•
42
Server / Storage agent option: SSLTLS12
•
Default is No (specifies that the server uses TLS 1.1 for SSL sessions)
•
YES (causes the server and storage agent to prefer TLS 1.2 )
•
Client, server & storage agent check that the session has been setup with protocol TLS 1.2 or higher
•
They will break the connection if TLS 1.2 or higher is not in use
SSLDISABLELEGACYTLS
•
Defaults to NO (SSL session processing works as in prior releases, client and server negotiate TLS level)
•
YES (SSL session setup is terminated if TLS 1.2 or higher is not used)
•
Overrides the SSLTLS12=NO option
•
Enforces the rejection of SSL connection attempts that use levels earlier than TLS 1.2
•
No compliance with certificates for FCM on VMware
•
Compliance for the CIMOM device adapter with FCM on Unix/Linux
•
Unencrypted TCP protocols are still allowed
Meet National Standards
TSM Operation Center - Conformance to NIST SP800-131a
•
Secure browser to OC connection
•
During Install of OC 7.1.1 option for which type of secure connection (or none at
all) that the browser will use to connect to OC
• Off – The connection will not be secure.
• Transition – The connection will use TLS 1.0
• Strict – The connection will use TLS 1.2
43
TSM 7.1.1 Compress Database Backups
Description: Selectively choose which TSM database backups are compressed
Benefits:
–
–
–
–
TSM deduplication increases size of the TSM database
The space requirements have also increased when backing up the TSM database
DB backup to local media: compression doesn't add much, compression overhead can be too much
DB backup to remote media: reduced amount of data sent compensates for compression’s overhead
Server Support:
– What can be configured
• Can specify a "compress" parameter on the BACKUP DB and SET DBRECOVERY commands
• SET DBRECOVERY sets the compress defaults for the BACKUP DB command
– Impact of configuration decisions
• Processor usage and duration increased
• Increases varied based on database size, and hardware configuration
• Database backup compression decreases the size of the backed up database files
Backup file sizes: For a brand new database “sample” (backed up and compressed shortly after DSMS
TSM 7.1.1 Compress Archive Logs
Description:
• Compress DB2 archive logs
• DB2 10 introduced the ability to compress log files as they are archived
• In TSM 7.1.1
• Default to compress archive log is set to “NO”
• Known DB2 active log problem that occurs when the DB2 archive log compression flag is set to YES
• Only seen the DB2 active log problem on our large appliance CET test systems when they are under
sustained high volume usage and heavy workloads
Benefits:
• Consume less space and reduce the time required to back up the archive logs
• Potentially reduces the frequency of back ups that are initiated just to clear the archive logs
Server Support:
• Default ARCHLOGCOMPress value is No (do not compress archive logs)
• ARCHLOGCOMPress option set in the dsmserv.opt file and by the SETOPT command
• dsmserv.opt: server must be started or restarted for the option to take effect
• SETOPT command: can be used when the server is up and running
• Updates the server option immediately and appends the option and value to the dsmserv.opt file
• Current setting of the server option displayed with:
• Q LOG F=D
• Q OPT
• Q OPT ARCHLOGCOMPress
45
TSM 7.1.1 Solaris x86 native 64-bit B/A Client
●
Solaris x86 (32-bit) package was last updated in TSM V6.2
●
Solaris x86 continues to be a popular customer request
●
7.1.1 updates the Solaris x86 build to the most current Backup-Archive client version
46
–
64-bit native package will be updated as well
–
Solaris 10 and 11 supported
Currency
TSM for Space Management 7.1.1
 TSM 7.1.1 GPFS User Exit based HSM FAILOVER
– Replaces legacy HSM failover method
– HSM uses GPFS user exit callback function to register for failover modes
– HSM uses GPFS user exit callback function to identify a cluster node as HSM node
– Simple functionality to improve servicability
– New failover modes to improve usability (Enabled, TakeOver, HandOver, Disabled)
– The scout daemon (if used) stores the port number in file
/etc/adsm/SpaceMan/config/DSMSCOUTPort
 Enhanced Recall provides more granular recall
– dsmmigfs, dsmmigrate, dsmattr commands enhaced with more granularity for recalls
– PREVIEWSIZE option added to prevent file explorer preview icons from causing recall
47
FastBack For Workstations 7.1.1 – Admin Console
Moving central admin console from TIP to Websphere Liberty Profile
•
•
User interface basically unchanged
•
Smaller package, around 80% smaller
•
Install in 10 minutes instead of an hour
•
Provides greater flexibility for future releases
Adding graphing to the reporting tables in Central admin console
•
•
Easier to understand data
•
Every report will have a button to display a graph of the data
•
When more than 50 rows are present, sliders appear on the graph for better usability
Support for Windows Servers 2012 and 2012 R2
•
48
FastBack For Workstations 7.1.1 – Client
New block level backup strategy
•
•
Results in fewer full backups, since block level is used mostly for very large files, like email
•
New block level will be fully backward compatible with no changes to the configuration
Help system will be upgraded and a link to the IBM Knowledge Center will be added
•
•
Product integrated help for the client is now created by using Eclipse tools
•
Runs on all modern browsers
Encryption has been strengthened to 256 bit

•
Files backed up with native file encryption may use 256 bit encryption
•
Set option by running command from install directory: fpa config-set Encryption Type=2
Support for Windows 8.1 and 2012 R2
•
49
Enablement Resources
Youtube short demos: http://www.youtube.com/watch?v=el_se-m8QoI&list=PL1c3vkFzmhec0xvPtplZD_yNpaPNfRHh0
– Tivoli Storage Manager Operation Center scheduling:
– Tivoli Storage Manager Operation Center storage pool and policies:
– Tivoli Storage Manager Operation Center Client Management Services:
– Tivoli Storage Manager Operation Center Advanced Filtering:
– Tivoli Storage Manager Operation Center Executive Email:
– Tivoli Storage Manager Node Replication dissimilar polices and damaged file recovery:
– Tivoli Storage Manager for VE 7.1.1 - Data Protection for Hyper-V - Protecting Hyper-V Virtual Machines
– Tivoli Storage Manager for VE 7.1.1 - Data Protection for Hyper-V - Individual File Restore
50
Thank You
51
|

similar documents