MODAPS PCR 21-027
| PCR Number | 21-027 |
|---|---|
| Date | 2021-10-03 |
| Initiator | Sudipta Sarkar |
| Abstract | PGE117 6.1.10, PGE118 6.1.10, PGE119 6.1.10 -- Fixed a bug of callocing QC_16A2_line_temp of 8-day under the nested loop of pixel of each line but it should be calloced under the loop of lines. Important note: The bug did not introduce any errors to the A2 outputs but took >14G memory for PGE118 or 8-day MxD16A2 processes. |
| Problem Statement | The MOD16 8-day process was consuming too much system memory, much more than expected limits, and was causing other competing processes to slow down. |
| Description of Change | The PGE117 that is the fundamental driver PGE was updated to correct a bug that was allocating memory for certain variables within a nested loop, thus leading to memory leakage and consuming more memory than needed by the process. |
| Products Affected | Daily C61 L3 regular and gap-filled MxD16 products. |
| Software Affected | PGE117 6.1.10, PGE118 6.1.10, PGE119 6.1.10 |
| Processing String to Receive the Change | Terra and Aqua Forward Processing. Terra and Aqua Reprocessing. NRT. |
| Downstream Product Effects | None. |
| Data Granules to be Used for Science Testing | A science test was conducted to confirm the successful generation of all C61 gap-filled MxD16 products in AS 1874 and verified by LDOPE. |
| Justification | The fix made in PGE117 V6.1.10 will allow for more efficient consumption of system memory and will minimize any leakage and also allow other competing processes to run in the production minions and will not cause any unnecessary delay because of higher than required memory usage. |
| Effective Date for Implementation of Change | C61 PGE117/118/119 V6.1.10 will be inducted for C61 re-processing and leading edge of forward processing starting 10/05/2021. |
| Status | Major algorithm development completed. Sadashiva Devadiga approved on 10/13/2021 Gang Ye approved on 10/29/2021. |
