Variation in reported times

[oj]

Hi,

When I run a particular query on Sparksee for n number of times I get the following result in milliseconds (n=7):

63

15

0

16

15

0

1) I understand that the first time, the time taken is high due to operating on a cold cache. But can you please explain why the fluctuating behaviour for the rest of the runs? This is just an example, I encounter similar behaviour for couple of others as well. I believe this is to do with the way the caching happens in Sparksee?

2) Also, if I want to run the same query with a different parameter (again 7 times), is there a way to clear the cache before running it?

[sparkseegdb]

Hello,

Hope we can be of help!

1) You're right, the first time is probably that the cache is empty. The other times are much more difficult to explain while the cache may have a big influence in these times it will also affect if you repeat the same query with the same arguments, if its a single session, if the query is read only or includes updates or even the recovery activated. 

2) The only method to completely clear the cache would be to close the database and open it again. But you should also completely clear the operating system file cache if you want an absolute cold start.

Could you tell us a little bit about what are your tests for? So we can understand the type of support you would need for this.

Regards,

Sparksee team