Support #12223
[SCP ID :##5958##] : Improving Text File Report Memory while writing large record
| Status: | Closed - End of life cycle | Start date: | August 20, 2021 | |
|---|---|---|---|---|
| Priority: | Normal | Due date: | ||
| Assignee: |  Zahir Abd Latif | % Done: | 100% | |
| Category: | - | Spent time: | - | |
| Target version: | - |
Description
Hi,
Kindly attend below request:-
We are hitting memory issue while running TAR02 report (txt report) for MBB in production. The report generation seem to be dead and did not proceed to generate remaining report. We found out that MBB transaction volume is hitting 3.9 million.
ITO attempt to rerun the report but still failed. However we later manage to rerun the batch job by allocating more memory heap size when running the job as temporary.(Previous running at 10G , change to run 20g)
Moving forward, management have request us to review current code and evaluate efficient way in writing to the report file. Can you help to review and suggest on how we can improve this.
Scenario criteria:
- Report TAR02 generation
- Transaction volume > 3.9 million
- Running job at 10G heap size allocation.
- Server Memory 64GB
History
#1
Updated by Ngoh Chee Onn about 3 years ago
- Status changed from New - Begin Life Cycle to Pending Customer Feedback
- % Done changed from 0 to 70
Hi All,
Alternative solution as below had been propose to customer and pending customer feedback.
1) Add in flush method on Output file writer to flush out the buffer memory in JVM.
2) Add pagination query on the transaction data by dynamic parameters records. For example, the total transaction to be written into file is 1000, pagination query will split the database query into 5 and 200 records by each call.
Thank you
#2
Updated by Ngoh Chee Onn about 3 years ago
- Assignee changed from Ngoh Chee Onn to Zahir Abd Latif
- % Done changed from 70 to 90
Hi Zahir,
Solution package has been delivered and pending customer testing on SIT environment.
Thank you
#3
Updated by Zahir Abd Latif about 3 years ago
- Status changed from Pending Customer Feedback to Closed - End of life cycle
- % Done changed from 90 to 100
Issue closed in SCP.