M_SERVICE_REPLICATION table fields

[Former Member]

Hi Team,

We are in the process of setting up alerts whenever DR is behind by some particular number of log files we would be reporting an alert. Now we have selected these two field to report the difference.

SHIPPED_LOG_POSITION

BIGINT

Shipped Log Positon

LAST_LOG_POSITION

BIGINT

Current Log Pos

The problem being is this we have given a condition where the difference between the two is greater than 100. What we wanted to understand is that what are the numbers indicated in the two fields. Is it a good indicator as to how we are achieving it.

We are getting frequent alerts where the log difference which is (last_log_position - shipped_log_position)  "log difference is 245184" or some particular number. What i wanted to understand what exactly is a good number to start with for the alerting?

[Former Member]

The log position is rather fine grained and not linked to the log files. I would recommend you to monitor the log shipping time delay instead, i.e. check columns SHIPPED_LOG_POSITION_TIME and LAST_LOG_POSITION_TIME. This approach is already used by the SAP HANA Mini Checks (SAP Note 1999993), Check 1815, threshold 60 seconds.

[Former Member]

Martin,

I was thinking about you and you answered .

I was checking this file HANA_Configuration_MiniChecks.txt.

Is this the one should i use?

[Former Member]

Martin,

Is this the statement you are referring to ?

SELECT

        'LOG_SHIPPING_DELAY',

        '',

        IFNULL(TO_CHAR(MAX(SECONDS_BETWEEN(SHIPPED_LOG_POSITION_TIME, LAST_LOG_POSITION_TIME))), '0')

      FROM

        DUMMY LEFT OUTER JOIN

        M_SERVICE_REPLICATION ON

          1 = 1

[Former Member]

Yes, this is the section. It should be the same for all variants of the Mini Checks. I basically check for:

MAX(SECONDS_BETWEEN(SHIPPED_LOG_POSITION_TIME, LAST_LOG_POSITION_TIME))

If it is larger than 60 (seconds), a "potentially critical" flag is raised.

[Former Member]

Why have you chosen as 60 seconds?

Isnt it a too low a value?

What is this number indicative about the database LOG_POSITION? Is it like an LSN in oracle?

[Former Member]

60 seconds is just a "best guess". All the systems that I have monitored so far haven't hit the 60 seconds limit, so I currently think that it is a good limit to be alerted.

LOG_POSITION is much more granular than LSN, you can think of it more on SCN level.

[Former Member]

Martin,

Probably a basic question, cant we get the host and secondary host fields as well will it possible when we use functions?

[Former Member]

SQL: "HANA_Replication_SystemReplication_Status" (SAP Note 1969700) provides both information about the hosts, the shipping delay and the SR status, e.g.:

-------------------------------------------------------------------------------------------------------

|PATH      |HOSTS                     |PORT |SHIP_DELAY_H|MODE   |STATUS|STATUS_DETAILS               |

-------------------------------------------------------------------------------------------------------

|ha1 -> ha2|saphana2100 -> saphana2101|30205|       27.46|SYNCMEM|ERROR |Log shipping timeout occurred|

-------------------------------------------------------------------------------------------------------

[mathanponnucham]

Hello,

Your time to recover and system availability SLA dictates the good number you are looking for to start with alerting.

Rgds,

Mat.

[Former Member]

Mathan,

Thanks for your response, can you please check my query again.

What are those numbers i still cannot relate to the database as to how big a gap is or can be?