Option to provide Batch size for each rule

@HollySimmons did you get a chance to view the comments here. 

Thanks for the nudge @sai_ram 

@rakesh this was the answer/suggestion from support; 
 

We actually reduce batch size for the instances which are on SFDC edition and not the NXT ones . As in NXT instance, the rule action uses  ' SFDC Bulk Api V2' (https://developer.salesforce.com/docs/atlas.en-us.api_bulk_v2.meta/api_bulk_v2/) for your tenant. This means we are no longer relying on the Apex Rest APIs from Gainsight managed package(used in SFDC edition instance) and instead using the API which SFDC has provided. In this case, the batching of records is handled by SFDC and we cannot reduce the batch size for rules.

In order to fix this, there are two options-

Either disable or adjust the triggers that are causing the issue or,
Exclude the Integration user(one who authorizes SFDC connection) from the SFDC triggers? 

But the options suggested aren’t feasible as the update we’re doing is in order to kick off the trigger itself. We know from our own testing using data loader that it can handle about 5 records at a time (we’ve managed more, but at 5 it never fails), and the API listed above creates batches of 10,000 records. 

What we’re hoping for is something like the dataloader settings where we can choose to override that to our own determined batch size (per action)

PS incase you’re wondering, what we’re doing is converting leads to contacts via a trigger.

In SFDC we have a duplication rule which means email must be unique across both leads and contacts, so if someone exists as a Person in Gainsight but a lead in SFDC we then need to convert them to a contact in SFDC. So what we do is check leads in SFDC for a match, and if there is we just set a boolean field to true on that lead record, which then triggers the process to convert them. 

Hi Holly, Bulk API route for Load to SFDC is the approach we want to take in NXT because of the massive performance improvements we gain here (for 1 Million records the performance improvement is ~1000%). One enhancement we are thinking of is if the number of records are small, use rest calls. In that, we can think of an enhancement you are proposing.

cc: @swaroop_badam 

Thanks for the nudge @sai_ram 

@rakesh this was the answer/suggestion from support; 
 

We actually reduce batch size for the instances which are on SFDC edition and not the NXT ones . As in NXT instance, the rule action uses  ' SFDC Bulk Api V2' (https://developer.salesforce.com/docs/atlas.en-us.api_bulk_v2.meta/api_bulk_v2/) for your tenant. This means we are no longer relying on the Apex Rest APIs from Gainsight managed package(used in SFDC edition instance) and instead using the API which SFDC has provided. In this case, the batching of records is handled by SFDC and we cannot reduce the batch size for rules.

In order to fix this, there are two options-

Either disable or adjust the triggers that are causing the issue or,
Exclude the Integration user(one who authorizes SFDC connection) from the SFDC triggers? 

 

But the options suggested aren’t feasible as the update we’re doing is in order to kick off the trigger itself. We know from our own testing using data loader that it can handle about 5 records at a time (we’ve managed more, but at 5 it never fails), and the API listed above creates batches of 10,000 records. 

What we’re hoping for is something like the dataloader settings where we can choose to override that to our own determined batch size (per action)

 

PS incase you’re wondering, what we’re doing is converting leads to contacts via a trigger.

In SFDC we have a duplication rule which means email must be unique across both leads and contacts, so if someone exists as a Person in Gainsight but a lead in SFDC we then need to convert them to a contact in SFDC. So what we do is check leads in SFDC for a match, and if there is we just set a boolean field to true on that lead record, which then triggers the process to convert them. 

Highly needed! Where are we on this? It looks like you wouldn’t be considering it? What’s the workaround?