PK Chunking: Fixing Salesforce Bulk API Timeouts скачать в хорошем качестве

PK Chunking: Fixing Salesforce Bulk API Timeouts

Is Salesforce Bulk API consistently timing out when you try to query tables with large recordsets? PK Chunking can be a savior when we’re faced with frustrating timeouts. So, what is PK Chunking anyway? PK stands for primary key -- in that it speeds up our query by creating an index of Salesforce objects fields. In Salesforce, the primary key is the object’s indexed record ID. Chunking means its splitting up that mammoth data set into manageable chunks which boosts our performance. It’s a handy tool to manage large datasets. Before we dive into the details of PK Chunking, take a moment to subscribe to our channel – Salesforce Explorer. We’re exploring the breadth of Salesforce one feature at a time so join us for this adventure. How Does PK Chunking Work? Under the hood, PK Chunking is performing subqueries to break up that large dataset. Each chunk is processed as a separate batch which counts towards your daily batch limit. It is including WHERE clauses on the starting and ending record IDs to define the successive batches within the bulk query. It is running these separate batch processes in parallel which significantly cuts down on job run times. How to Enable PK Chunking? Enabling PK Chunking in our Bulk API query job request is straight forward. In the header, we simply include the line: Sforce-Enable-PKChunking There are three parameters we can add – chunkSize, startRow and parent. By default, the chunk size is set at 100,000 records, but we can modify the chunk size by setting the chunkSize value. For example: Sforce-Enable-PKChunking: chunkSize=250000; Notice, 250,000 is our upper bounds. Increasing batch size will result into fewer batches, but they will be less performant. We also have the option of defining which row we start processing by using the parameter startRow. Sforce-Enable-PKChunking: chunkSize=250000; startRow=00130000000xEftMGH Our last one is parent field. This specifies which parent object we’re using when running queries on shared or history objects. The chunking will be based on the parent object’s record versus the shared or history object. For example, if we were looking at AccountHistory, we’d reference the Account object instead. Sforce-Enable-PKChunking: chunkSize=250000; startRow=00130000000xEftMGH; parent=Account What Happens When the Batch Runs? When this process begins, it will subdivide that original query into multiple batches. Once those batches have been queued up, the original batch status will update to NOT_PROCESSED. This status signifies that all batches have been successfully created. If the chunking process fails, the status will update to FAILED, but that doesn’t necessarily mean no chunks were created. Any batches that were successfully generated will process as expected. Lastly, when a job is successfully batched, you’ll need to track the progress through those individual jobs. The original batch will continue to show NOT_PROCESSED even when all batch jobs have successfully run. Helpful Tips Regarding PK Chunking • Salesforce recommends enabling PK chunking for objects that exceed 10 million records or when bulk query consistently times out. • PK chunking is supported for custom objects and most standard objects. I’ve included a link in the description below detailing which standard objects are supported. • Be careful when filtering by any field by ID or using an ORDER BY clause. Using these will effectively disable your PK chunking for the bulk query job. • When you define a chunk’s ID boundaries, soft delete records will be included in the results. They won’t process, but the returned result will be less than the defined chunk size. If you are struggling with tough problems in Salesforce, consider reaching out to the company I work for Improving. We have a team of awesome Salesforce consultants that can help. Please reach out to us at http://improving.com to explore how we can assist your team. We look forward to talking with you. Thank you for taking the time to explore PK Chunking with us today. I hope you found this video valuable. If so, please like and subscribe to our channel. We are publishing videos on interesting Salesforce topics several times a month, so we’d love for you to join us for our future sessions. We look forward to seeing you then. Standard objects that support PK Chunking: https://developer.salesforce.com/docs...