Hi Dinesh,
The AWTLimit field in ResUsageSAWT was first introduced in 14.0. AWTLimit did not exist in the SAWT base table in earlier releases, such as 12.0. These are the only AMP worker task related fields carried in ResUsageSAWT in 12.0:
MailBoxDepth
FlowControlled
FlowCtlCnt
WorkTypeInuse00 - WorkTypeInuse15
WorkTypeMax00 - WorkTypeMax15
I can only suggest that the PDCR routines calculated the AWTLimit field in the 12.0 system, since you are showing me output from the PDCR database, not DBC. You are comparing DBC tables in 14.10 against PDCR tables in 12.0.
You will have to speak with someone knowledgable with PDCR conventions to explain how they make that calculation, if you think it is incorrect.
One thing I noticed is that the 14.10 SAWT output looks like it is in summary mode, although I cannot tell for sure without seeing if there is an 'S' in the SummaryFlag colummn. The AWTLimit values are higher than is possible (2880) for normal ResUsageSAWT output. Divide 2880 by 80 (the default number of AWTs) and you get 36, which I assume is the number of AMPs on the node.
Summary mode rolls up all the SAWT AMP-level output to one row per node, instead of one row per AMP. Your WorkTypeInuse00 values in the 14.10 output are way too high as well for standard logging (254, 172, etc.). That also points to logging in summary mode.
The PDCR output on the other hand shows AWTLimit (calculated by PDCR) as 80, which is the usual default of AWTs per AMP. So I would guess that the output from 12.0 IS NOT in summary mode and the 14.10 output IS in summary mode.
Check the SummaryFlag value for both to validate.
Thanks, -Carrie
Hi Dinesh,
The AWTLimit field in ResUsageSAWT was first introduced in 14.0. AWTLimit did not exist in the SAWT base table in earlier releases, such as 12.0. These are the only AMP worker task related fields carried in ResUsageSAWT in 12.0:
MailBoxDepth
FlowControlled
FlowCtlCnt
WorkTypeInuse00 - WorkTypeInuse15
WorkTypeMax00 - WorkTypeMax15
I can only suggest that the PDCR routines calculated the AWTLimit field in the 12.0 system, since you are showing me output from the PDCR database, not DBC. You are comparing DBC tables in 14.10 against PDCR tables in 12.0.
You will have to speak with someone knowledgable with PDCR conventions to explain how they make that calculation, if you think it is incorrect.
One thing I noticed is that the 14.10 SAWT output looks like it is in summary mode, although I cannot tell for sure without seeing if there is an 'S' in the SummaryFlag colummn. The AWTLimit values are higher than is possible (2880) for normal ResUsageSAWT output. Divide 2880 by 80 (the default number of AWTs) and you get 36, which I assume is the number of AMPs on the node.
Summary mode rolls up all the SAWT AMP-level output to one row per node, instead of one row per AMP. Your WorkTypeInuse00 values in the 14.10 output are way too high as well for standard logging (254, 172, etc.). That also points to logging in summary mode.
The PDCR output on the other hand shows AWTLimit (calculated by PDCR) as 80, which is the usual default of AWTs per AMP. So I would guess that the output from 12.0 IS NOT in summary mode and the 14.10 output IS in summary mode.
Check the SummaryFlag value for both to validate.
Thanks, -Carrie