HANA Replication - Alert "Internal event" (id 21)

[Former Member]

Hi

I am receiving this email alert:

snapshot   : 2015-08-18 16:07:52.0000000

Alert Name : Internal event

Rating     : Error

Details    : ReplicationError with state INFO with event ID  999 occurred at 2015-08-09 17:14:17.0000000 on hostp01:30101. Additional info: parameter mismatch (existence): indexserver.ini/database/hp1/[sqlscript]/enable_select_into_scalar_udf = 'true' exists only on site 1

User Action: Resolve the event and then mark it as resolved by executing the SQL statement ALTER SYSTEM SET EVENT HANDLED '<host>:<port>' <id>.

I updated this parameter on the replication site to match the primary, and would like to mark this event as resolved, but getting the following:

ALTER SYSTEM SET EVENT HANDLED 'hostp01:30101' 999 (on tenant db)

Could not execute 'ALTER SYSTEM SET EVENT HANDLED 'hostp01:30101' 999' in 1 ms 190 µs .

SAP DBTech JDBC: [7]: feature not supported: Can only process events for own service.

ALTER SYSTEM SET EVENT HANDLED 'hostp01:30101' 999  (on system db)

[129]: transaction rolled back by an internal error: exception 2150004: State 4 not allowed for info type

Any advice?

Kind Regards

Willem

[Former Member · Accepted solution]

Handleing of replication alerts doesnt worked in my situation too.

As these alerts are known to me. So,I tought dissabling parameter inifile_checker in HANA will work in this case. This will eliminate generating the alerts and checking the ini files between primary and secondary

Alternatevely, it is possible to disable check for a specified entry in INI file using the following syntax under global.ini/[inifile_checker].

exclusion_[inifile name|*][/<LAYER>] = [section with wildcards|*][/parameter with wildcards|*], ...<LAYER> := SYSTEM\|HOST\|TENANT

[Former Member · Accepted solution]

Hi Pavan

Thanks, Very helpful.

Yes, after the parameter was matched with the primary, I continued to receive the alerts.

The parameter I matched in the indexserver.ini, was:

[sqlscript]

enable_select_into_scalar_udf = TRUE

I added this parameter to inifile_checker to disable checks (under global.ini): 

exclusion_indexserver.ini/system = sqlscript/*  (is this the format?)

Kind Regards

Willem

[Former Member · Accepted solution]

Yes, format looks good.

If the format doesnt work, as this parmater is related to repliation, you can safely try another format. But, for me it looks good.

[Former Member · Accepted solution]

Hi Willem,

The mismatch parameters when updated on the seondary systems need to be followed up with the below command to ensure the system is aware of the new parameter

hdbnsutil -reconfig

I have used this step many times and it works

I would not recommend making changes to INIFILE CHECKER as going forward in SPS09 we have multiple new .ini files coming up which do not exist by default in INIFILE CHECKER and you may have to keep creating them each time

Hope it helps

Sunil

[Former Member · Accepted solution]

Hi Sunil,

Thanks, I have now Executed "hdbnsutil -reconfig" on both the Secondary and Primary sites (after I matched the ini files) - and I still get the alert.  Will still look for ways to resolve this alert.

For now, Disabling the checks using inifile_checker parameter works to stop the emails (I notice I needed to add the parameter to the SYSTEMDB, adding  it to the TENANTDB did not stop the emails, even from TENANTDB alerts)

Kind Regards

Willem

[Former Member · Accepted solution]

snapshot   : 2015-12-28 10:33:55.0000000

Alert Name : Restarted services

Rating     : Error

Details    : indexserver on host <XYZ>port 30003 was restarted at 2000-01-01 00:00:00.0000000.

User Action: Investigate why the service had to restart or be restarted, for example, by checking the service's trace files.

snapshot   : 2015-12-28 10:32:38.0000000

Alert Name : Restarted services

Rating     : Error

Details    : compileserver on host <XYZ>port 30010 was restarted at 2000-01-01 00:00:00.0000000.

  User Action: Investigate why the service had to restart or be restarted, for example, by checking the service's trace

snapshot   : 2015-12-27 22:31:55.0000000

Alert Name : Restarted services

Rating     : Error

Details    : webdispatcher on host <XYZ> port 30006 was restarted at 2000-01-01 00:00:00.0000000.

User Action: Investigate why the service had to restart or be restarted, for example, by checking the service's trace files.

So for these i have update the inifile in the systemDB.

Parameters would be :

exclusion_nameserver.ini/system = landscape/*, statisticsserver/*

exclusion_global.ini/system = storage/*, persistence/*path*, *hostname_resolution*, system_replication_site_masters/*, system_replication/*, memorymanager/*, inifile_checker/*

[Former Member]

Hello,

Disabaling checks is a workaround.

Check note 2253869 you have to set the event acknoledged, then deleted and finaly handeled.

Regards

[Former Member]

Hi Willem,

Have you got any more alerts after the parameter was matched with the primary?

The ALTER SYSTEM SET EVENT HANDLED some how does not work and gives the same error for me

Sunil