Product Version

6.5 Update 5

Incident Reference: 

INC0011720

Problem Statement: 

Replication records stuck/ not moving from HO to all available stores

Details: 

Replication records were stuck at HO and piled up for each and every store and they were not moving at all

Proposed Solution/ Workaround:

We found out that few stores were not in use and because of these stores replication of other stores were stuck, on taking required approval we marked those stores as deleted and cleared the senddump table

Steps:

Step 1 -

On executing the given query (present in screenshot) at HO, we found that last sequence no. of few stores (highlighted in yellow) were completely different to rest of the stores.

It means that these stores were either no more doing any transactions or have marked as deleted.

Step 2 -

We have to confirm that if these sites have been marked as deleted by running the query below, and checking 'isdeleted' column, which is having value as '0' that is equal to 'not deleted'

Step 3 -

Step 4 -

After all above is confirmed and checked, we proceed to resolve the issue.

a.) Required approval is taken on session chat/ email to have the stores deleted that are no more active/ and have any active transactions happening 

b.) Stop the replication service at HO

c.) Run the update query as present in screenshot below

d.) Run 'delete from RepReplicationSendDump' query

e.) Start the replication service stopped earlier

Step 5 -

On starting the replication services, to check if the issue is resolved, please retrieve siteid of any of the stores (excluding deleted stores), 

and run the first query, and hit it in interval of few minutes, if the lastsequenceno is increasing then the replication has started moving from HO to this store, and issue is resolved

​

Constraints/ Scenarios: 

1. Please do not execute any update/ delete queries if scenario is different to the one stated above.