Throughput

Database
Enthusiast

Throughput

Can anyone tell me the diff b/w throughtput and response time
and if possible ans the below queries

What distinguishes throughput from response time?

a.Throughput measures how many rows are processed. Response time measures how long a process takes.
b.Response time measures how many queries are processed. Throughput is how long the processing takes.
c.Throughput measures how many queries are processed. Response time is how long the processing takes.
d.Response time is how long the processing takes. Throughput measures how long the rows are processed.

What could improve response time but could also decrease throughput?

a.decrease priority of the workgroup
b.decrease the number of concurrent users
c.increase the number of concurrent users
d.increase spool space for concurrent users
5 REPLIES
Enthusiast

Re: Throughput

Through-put typically has to do with the number of queries processed in a given period of time. Can be expressed as queries per second or minute. Response time is the time it takes for a given query/transaction/workload to complete.

With that definition, if we look at the next question, we'll see that option a probably has a negative impact on both response time and throughput. Fewer resources allocated to a workgroup mean longer response times and probably lower throughput as queries take longer to complete meaning subsequent queries are delayed as well.

Increasing the number of concurrent users would increase throughput but could potentially elongate response times as more queries compete for the same resources. Increasing spool space probably has little effect on response time or throughput, although it could allow more queries to run at once which would increase throughput.

So that leaves us with option (b). By decreasing the number of concurrent users, there are less queries competing for resources at any point in time. This means they should finish faster which gives us a better response time. But fewer concurrent users means that we'll also have fewer queries executed in the same amount of time which leads to a lower throughput.

Of course, it's really a balancing act. If we decrease the concurrent users, queries should finish faster. Assuming a constant flow of queries, this means that each user will submit more queries in the same time than they would have with slower response times. Which means the throughput is higher than if we had reduced concurrent users but seen no impact to response times.
Enthusiast

Re: Throughput

we are using TD 13 and for few tables we are getting the ThroughPut/Min as 0.3 GB. We are trying to load the table using tpt load operator, using named pipe as input file from AIX server, Can someone let me know what are the ways of increasing the Throughput/Min

Enthusiast

Re: Throughput

I feel c and b.

Cheers,

Raja

Enthusiast

Re: Throughput

Hi Bobby ,

your questions are very useful to practise..Can you please share some more questions on Terdata for practise

Advance Thanks,

Sarani

Enthusiast

Re: Throughput

Hi Raja & Others,

I red all your posts Could you please help me on Teradata practise questions and how to crack Teradata 14 basic.

Advance Thanks,

Sarani