HANA Multi-host Replication Certificates

[Former Member]

Friends!

Hana 3 host system version 97
SUSE linux server 11 sp3

I create a system with same topology version 10

Add new system as replication slave,

replication mode: sync

1.) Do Certificates replicate over?

2.) Does all XS engine Content replicate over?

I have been stuck in some old sap cockpit, webdispatcher ui problems during recovery and update process.

I have been updating my hana system and then Recovering and that order of operations is overwriting the updates in xs engine administration UI.

Would replicating over write my xs engine administration ui? /sap/hana/xs/admin OR /sap/xs/admin/webdisptacher/admin ?

Thank you for your time and diligence!

[prashb · Accepted solution]

Hello Zachary,

This is what i understand from your post, you have setup system replication between :

primary system:

Hana 3 host system version 97 (Revision)

SUSE linux server 11 sp3

secondary system:

Hana 3 host system version  10 (SPS)

SUSE linux server 11 sp3

Replication mode: sync  primary -> secondary  . you should not do secondary -> primary replication, since secondary is higher version than primary. 

1.) Do Certificates replicate over?

certificate stored in the database will be replicated but not those stored in the file system.

Certificate Management in SAP HANA - SAP HANA Security Guide - SAP Library

2.) Does all XS engine Content replicate over?

yes, XS engine content will also be replicated.

Regards

Prashanth

[Former Member · Accepted solution]

I am also having trouble understanding adding hosts and system replication..

3 hosts on the primary Hana version 97
streaming to
3 host replication server; Hana version 100

IF the secondary must have equal number as the Primary how does one add hosts?

I am planning on using secondary for minimal down time.

Snapshot and upgrade original primary, add host, distribute data.

Then what?

1.) We have an app that will sync data to primary so it is "Current"

or
2.) any more documentation on hdbnsutil and adding hosts?

[prashb · Accepted solution]

Yes, the number of nodes in the secondary system has to be equal to the number of active nodes in the primary system - so you can only add standby nodes.

below link takes you to admin guide section "Use System Replication for Near Zero Downtime Upgrades" - this explains how you can use secondary for minimal or zero down time maintenance of primary system:

SAP HANA Administration Guide - SAP Library

hdbnsutil  -  is the command line tool which helps you to setup and work with system replication.

Set up System Replication with hdbnsutil - SAP HANA Administration Guide - SAP Library

to add host you have to use resident hdblcm or hdblcmgui or hdblcm web

Adding Hosts to an SAP HANA System - SAP HANA Administration Guide - SAP Library

[Former Member · Accepted solution]

I performed the fail over process;

Shot the Primary in the head,

run sr_failover on Secondary
Moved Virtual IP to Secondary,

Success.

Moved Primary(Original Primary) to new Virtual Private Cloud AWS(Snap shot, launch instance from snap shot in new VPC; Opened all Ports
Updated version 100
Ran Data Sync from third party resource; to "Catch up" , primary and secondary have same data,

global.ini communication listening .global

Shot Secondary(Current Primary) in the head,

Moved Virtual IP to Primary(Original Primary),
Problems;
Hana studio Administration Console showing no data, only maintains REFRESHing state.

CAN see tables, CAN query data.

No cockpit administration, no xs admin

all services running but something was funny on HDB info

usually

\_ /usr/sap/H2D/HDB00/imdbmasterslave/trace/hdb.sapH2D_HDB00 -d -nw -f /usr/sap

h2dadm     3551   3324  0.2 13022284 1597316      \_ hdbnameserver

h2dadm     3963   3324  0.1 2891192 435968      \_ hdbcompileserver

h2dadm     3967   3324  0.1 2817276 453060      \_ hdbpreprocessor

h2dadm     3991   3324  2.9 55411612 52010500      \_ hdbindexserver

h2dadm     3993   3324  0.2 4318588 1175408      \_ hdbscriptserver

h2dadm     3995   3324  2.4 10413092 7480128      \_ hdbxsengine

h2dadm     4693   3324  0.1 3203748 518232      \_ hdbwebdispatcher

But the server in question had the hdbindexserver branching from hdbcompileserver?

Moved Virtual IP back to Seconday.

Attempted process again with Original primary before the snap shot
no success

Current state:

Secondary(Current Primary)
Primary( Original, after snap shot in AWS VPC)
Success.

////////////////////////

So my question is: Is this probably a networking issue? Hana internal routing holding the original private IP used for internal communication during hana install?

I updated /etc/hosts with proper internal ip.
hostname is also the same.

Updated:
global.ini
system_replication_hostname_resolution
Tried; sr_unregister -force
Tried; sr_disable
ERROR: Host has registered slave

Any ideas would be appreciated,
Thank you for your diligence,
Zachary.

[Former Member]

Hi,

After setting up the SYSTEM Replication you will notice the list of volumes that are replicated from TIER1 to TIER2 in the LANDSCAPE-->SYSTEM REPLICATION tab

The data contained in all these volumes is replicated and available on the TIER2, that would give you a better picture of the System Replication services

Hope this helps

Sunil

[Former Member]

Hana studio Administration view would not finish.. blank and in refreshing state.

thank you for your time.

[prashb]

hana studio may hang when you have system replication setup with mode sync and network connectivity is lost between primary & secondary.

Synchronous with full sync option (mode=sync. Full sync is configured with the parameter [system_replication]/enable_full_sync) means that log write is successful when the log buffer has been written to the log file of the primary and the secondary instance. In addition, when the secondary system is disconnected (for example, because of network failure) the primary systems suspends transaction processing until the connection to the secondary system is re-established. No data loss occurs in this scenario.

use the below url for correct order of executing hdbnsutil.

In our landscape we are using "async" replication with parameter logshipping_async_wait_on_buffer_full to false, we still have issue with hanastudio when we simulated network connection lost between primary and secondary. We have opened OSS message and SAP has accepted this is bug and they are working on it. we are at Rev 97.03.

Regards

Prashanth

[Former Member]

That was helpful thank you..
He did not mention stopping the primary..
I therefore jumped to conclusions..
Thank you.

The primary can continue to run and it will reconfigure itself after take over is complete, then I can take it off line to add hosts and distribute data and catch up to the secondary.. then do the switch since it cannot be a secondary, as it will not have equal number host.

Cheers!