SAP HANA Mission Critical Data Center Operations

SAP HANA Product Management December, 2015
(Delta from SPS 10 to SPS 11)
SAP HANA SPS 11 - What’s New?
Mission Critical Data Center Operations
Scale-Out, High Availability and Disaster Recovery

© 2015 SAP SE or an SAP affiliate company. All rights reserved. 2Confidential
Mission Critical Data Center Operation – What’s New in SPS 11?
Agenda
Scale-Out, High Availability (Host Auto-
Failover)
 Operational Security: Time-check during integration
of new Scale-Out members
Disaster Recovery, High Availability
 SAP HANA Storage Replication
o Changes in certification management
 SAP HANA System Replication
o New option: HotStandby (continuous Log Replay)
o Snapshot backups for starting SAP HANA System
Replication initialization
o ProxyViews of Secondary on Primary with SAP
HANA System Replication
o HA/DR Provider Framework
o New Hook: srConnectionChanged()
o Miscellaneous
o Good to know…

What’s New in SAP HANA SPS 11: Scale-Out/Host Auto-Failover
Operational security for Scale-Out
Operational security:
For new members of a Scale-Out cluster a check for local time settings
during integration of this new host(s) is performed
Standard is to prevent an integration of new Scale-Out members with
bigger time differences
Alternatively the installation tools (hdbaddhost) got manual overwrite
options for the graphical and console tool versions
 hdbaddhost --ignore=check_timediff
Additionally there exists an alert which checks regularly on time
differences in Scale-Out clusters
 Time difference on a host/member to initiate alert: 1,2,3 minute differences
equal to Low, Med, High priority warning
 Frequency: every hour
 Alert number: 76 – Discrepancy between host server times

Disaster Recovery
High Availability

What’s New in SAP HANA SPS 11: SAP HANA Storage Replication
Changes in certification management
Existing storage replication solution certifications for all
SAP HANA appliances continue their validity
Solutions reported in SAP note 1755396
All newer solutions are supported directly by corresponding
vendors (hardware and/or storage partners)
No further certification of these storage replication solutions
by SAP for use with SAP HANA
With the huge pool of certified technologies by SAP ICC for
SAP HANA tailored data center integration (TDI), SAP HANA
can use a lot of option of these hardware and technology
vendors
Data Center 2Data Center 1
OS: Mounts
Data
Volumes
Log
Volume
OS: DNS, virt. hostnames, virt. IP
Primary
Name
Server
Index
server
Name
Server
Index
server
Name
Server
Index
server
Secondary
Prod. (inactive), QA&DEV (active)
Name
Server
Index
server
Name
Server
Index
server
Name
Server
Index
server
HASolutionPartner
Storage
Mirroring
Clients Application Servers
HASolutionPartner
Data
Volumes
Log
Volume
Data
Volumes
Log
Volume
Data
Volumes
Log
Volume
Data
Volumes
Log
Volume
Data
Volumes
Log
Volume

What’s New in SAP HANA SPS 11: SAP HANA System Replication
HotStandby – Continuous Log Replay - Overview
Continuous log replay (Pure log-based transfer)
Redo Log is processed immediately on Secondary (Commit based)
New Operation mode “logreplay”
 Option to hdbnsutil: --operationMode=logreplay
 Has to be defined on Secondary connecting to Primary
No more delta-data transfer necessary
Reduced take-over times
Network traffic significantly reduced
Automatic initial data transfer still supported
Operations mode delta data shipment still supported
 Option to hdbnsutil: --operationMode=delta_datashipping
Compatible to the current set of operation commands – optimization under the hood
o Parameter name operation_mode [delta_datashipping|logreplay]
o Multi-Tier system replication supported
o No mixture with delta-data shipment (all members must use same operation mode)
Build the foundation for active/active operations (planned beyond)
Automatic Initial
Data Load
Primary
HANA
Secondary
HANA
Log
Delta-
Data
Primary
HANA
Secondary
HANA
Client
r/or/w
Log

HotStandby – Continuous Log Replay – Activation
Only a few steps are necessary to configure
SAP HANA system replication:
 Create an initial data backup (native or storage snapshot)
 Enable primary system for system replication
 Register secondary
 New option in Operation Mode: “logreplay”
 Start secondary
Command line tool hdbnsutil
First initial data backup
Enable Primary
 hdbnsutil -sr_enable ...
Jump to Secondary and register
 hdbnsutil –sr_register
–-replicationMode=[sync|syncmem|async]
--operationMode=[delta_datashipping|logreplay]...
Start Secondary
SAP HANA studio
Prim.
Sec.

HotStandby – Continuous Log Replay – New Parameters
logshipping_max_retention_size <> 0 (soft limit)
 In disconnect situation the log segments are not re-used until max size limit
is reached or in log full situation – then they are reused.
 Prevents system standstill on primary site due to too many log segments,
that are held for syncing the secondary site.
 This way primary is kept running with the drawback, that the secondary
cannot sync anymore and will have to be reinitialized with a data shipping.
logshipping_max_retention_size = 0 (hard limit)
 log segments required for secondary syncing are not re-used and a log full
results into a system standstill on primary site until log writing can continue.
 Here to sync the secondary has priority over standstill on the primary.
 When the reason for the log full has been resolved (on primary or secondary
site), transaction processing can continue.
 Consequence: Increase Log Volume capacity in case of longer
outages of system replication to prevent following initial loads.
 Alternatively: De-registration
A re-registration will automatically check if an initial shipment is again
necessary and execute accordingly (e.g. on longer extractions)
New
New

HotStandby – Continuous Log Replay – Feedbacks from Beta Shipment program
Comparison between
delta_datashipping and logreplay
with regards to network traffic showing the
expected
o (measurements done with itop Linux tool capturing
network traffic every 5 minutes):
 Delta data shipping: peaks roughly every 10
minutes (when delta data shipping is triggered)
 Logreplay: no peaks – continuously shipped log
buffers

HotStandby – Continuous Log Replay – Operation
Dual purpose use – running non-PROD on Secondary
Worst case memory requirement for shadow PROD operation: Persistent data – LOB data
 „Persistent data“ is all relevant content from data volumes loaded in memory
 „LOB data“ are large (binary|character) objects which are usually kept on disk (hybrid LOBs)
 Presumably half of Secondary occupied with shadow PROD operation
First sizing formula to evaluate situation from live systems:
Size of loaded column tables (in-memory)
+ RowStore size
+ 50 GB (for log replay operation)
Size of loaded column tables:
 select round(sum(memory_size_in_total)/1024/1024/1024) size_GB from m_cs_tables;
RowStore size:
 select host,round(sum(page_size * USED_BLOCK_COUNT)/1024/1024/1024,2) as
"RowStore Size GB"
from m_data_volume_page_statistics
where page_sizeclass = '16k-RowStore' group by host;
Further refinement probably follow…
Automatic Initial
Data Load
Primary
HANA
PROD
Secondary
HANA
shadow
PROD
HANA
non-
PROD
Log
Sizing?
Free
Resources

Snapshot backups for starting SAP HANA System Replication initialization
For starting SAP HANA System Replication an initial data
backup is necessary with a subsequent entry in SAP HANA
backup history
With SAP HANA SPS11 also a Storage Snapshot backup
listed in the backup history of SAP HANA can be accepted
Since SAP HANA Studio version 2.5 ff. this option is possible
in GUI tool
On console operation of SAP HANA System Replication it’s
already accepted today
 Benefit for administrators:
No waiting time for initial data backup to finish
SiteA

Renamed parameter
Renamed option for hdbnsutil:
--mode=[sync|syncmem|async]
has been renamed to
--replicationMode=[sync|syncmem|async]
Downwards compatibility is given
 Option --mode still recognized and interpreted as --replicationMode

ProxyViews of Secondary on Primary with SAP HANA System Replication
ProxyViews
For easier monitoring and reporting of the secondary site statistics, like
 detailed memory consumption,
 I/O-load and more
remote access via SQL is introduced on the primary site.
This feature is realized by providing proxy schemas and views on the primary site, which
extract the corresponding information from the monitoring views on the secondary site.
A new schema is available on the primary site named
_SYS_SR_SITE_<siteName>
where <siteName> is equivalent to the case-sensitive name used to register the secondary
site.
Note:
 Views can have an additional "<viewname>_RESET" view, e.g., "M_HEAP_MEMORY_RESET“
 Allows to reset the counters for a monitoring session on the selected view
Limitations
 Monitoring view access only possible, if primary and secondary run with exactly the same software version

HA/DR Provider Framework – new hook: srConnectionChanged()
HA/DR Provider Framework now offers an active
“communication” channel
 Possibility to inform external entities about activities
inside of SAP HANA operations
 New: Communication with an external cluster
manager to inform about trouble on the connection
to Secondary
 Detailed documented in SAP HANAAdmin Guide with
detailed code examples
 A Python script provides hooks that can be called in
response to events during System Replication process or
state change (see table)
 Parameter section in global.ini:
[ha_dr_provider_<classname>]
o provider = : The class name
o path = : Location of the script
o execution_order = : The ordering of the HA/DR Provider
if there is more than one; this is a number between 1 and 99
Name Trigger Caller Landscape Error behavior
startup() Beginning of
nameserver’s
startup phase
Starting
nameserver
Each host
individually
Nameserver
aborts, startup
is cancelled
shutdown() Just before the
nameserver exists
Stopping
nameserver
Each host
individually
Error trace is
written
preTakeover()
[System Replication]
s soon as the
hdbnsutil -
sr_takeover
command is issued
Master
nameserver
Called only
once on the
master
Takeover is
aborted
postTakeover()
As soon as all
services with a
volume return from
their assign-call
(open SQL port)
Master
nameserver
Called only
once on the
master
Error trace is
written
srConnectionChanged()
HA/DR Provider
hook for system
replication
connectivity
changes: As soon
as one of the
replicating services
loses or (re-)
establishes the
system replication
connection
Master
nameserver
Called only
once on the
master
Error trace is
writtenNew

Miscellaneous – Good to know…
Take-Over Decisions
SAP Note 2063657 gives some guideance on „Takeover Decisions“
 What helps when
Troubleshooting Guide
http://help.sap.com/hana/SAP_HANA_Troubleshooting_and_Performance_Analysis_Guide_en.pdf
SAP HANA System Replication for System Copy
Documented in SAP Note 2165547
Changes in the area of Alerts for SAP HANA System Replication
Alert 78 (Connection between systems in system replication setup – connection close) and
Alert 79 (Configuration consistency of system in system replication setups)
are now longer visible than 5 minutes  rechecked every hour

© 2015 SAP SE or an SAP affiliate company. All rights reserved.
No part of this publication may be reproduced or transmitted in any form or for any purpose without the express permission of SAP SE or an SAP affiliate company.
SAP and other SAP products and services mentioned herein as well as their respective logos are trademarks or registered trademarks of SAP SE (or an SAP affiliate
company) in Germany and other countries. Please see http://global12.sap.com/corporate-en/legal/copyright/index.epx for additional trademark information and notices.
Some software products marketed by SAP SE and its distributors contain proprietary software components of other software vendors.
National product specifications may vary.
These materials are provided by SAP SE or an SAP affiliate company for informational purposes only, without representation or warranty of any kind, and SAP SE or its
affiliated companies shall not be liable for errors or omissions with respect to the materials. The only warranties for SAP SE or SAP affiliate company products and
services are those that are set forth in the express warranty statements accompanying such products and services, if any. Nothing herein should be construed as
constituting an additional warranty.
In particular, SAP SE or its affiliated companies have no obligation to pursue any course of business outlined in this document or any related presentation, or to develop
or release any functionality mentioned therein. This document, or any related presentation, and SAP SE’s or its affiliated companies’ strategy and possible future
developments, products, and/or platform directions and functionality are all subject to change and may be changed by SAP SE or its affiliated companies at any time
for any reason without notice. The information in this document is not a commitment, promise, or legal obligation to deliver any material, code, or functionality. All forward-
looking statements are subject to various risks and uncertainties that could cause actual results to differ materially from expectations. Readers are cautioned not to place
undue reliance on these forward-looking statements, which speak only as of their dates, and they should not be relied upon in making purchasing decisions.

SAP HANA Mission Critical Data Center Operations

Empfohlen

Empfohlen

Weitere ähnliche Inhalte

Was ist angesagt?

Was ist angesagt? (20)

Andere mochten auch

Andere mochten auch (17)

Ähnlich wie SAP HANA Mission Critical Data Center Operations

Ähnlich wie SAP HANA Mission Critical Data Center Operations (20)

Mehr von SAP Technology

Mehr von SAP Technology (20)

Kürzlich hochgeladen

Kürzlich hochgeladen (20)

SAP HANA Mission Critical Data Center Operations