Re: [GENERAL] Predicting query runtime

2016-09-14 Thread Oleg Ivanov
Hi Vinicius, I recommend you to read this (http://www.doc.ic.ac.uk/~nb605/IO%20performance%20modeling%20research/Learning-based%20Query%20Performance%20-%202011.pdf) paper. Authors make a nice classification of different query performance prediction methods and propose their own solution for

Re: [GENERAL] Predicting query runtime

2016-09-13 Thread Vinicius Segalin
2016-09-13 17:12 GMT-03:00 Oleg Bartunov : > > > FYI, please check https://pgconf.ru/2016/89977 > > Interesting! Was this presentation filmed? Or would you have a post or something else with more details?

Re: [GENERAL] Predicting query runtime

2016-09-13 Thread Oleg Bartunov
On Tue, Sep 13, 2016 at 2:54 PM, Istvan Soos wrote: > On Tue, Sep 13, 2016 at 2:06 AM, Vinicius Segalin > wrote: >> 2016-09-12 18:22 GMT-03:00 Istvan Soos : >>> At Heap we have non-trivial complexity in our analytical queries,

Re: [GENERAL] Predicting query runtime

2016-09-13 Thread Istvan Soos
On Tue, Sep 13, 2016 at 2:06 AM, Vinicius Segalin wrote: > 2016-09-12 18:22 GMT-03:00 Istvan Soos : >> At Heap we have non-trivial complexity in our analytical queries, and >> some of them can take a long time to complete. We did analyze features >>

Re: [GENERAL] Predicting query runtime

2016-09-13 Thread Hu, Patricia
I’ve been looking for this on postgres too. Does Postgres have something similar to Oracle’s v$session_longops? It gives info on total unit of work, units done so far, last update time, and time remaining etc, and I found it valuable in providing an estimate to how long a certain query would

Re: [GENERAL] Predicting query runtime

2016-09-12 Thread Vinicius Segalin
2016-09-12 18:22 GMT-03:00 Istvan Soos : > Hi Vinicius, > > At Heap we have non-trivial complexity in our analytical queries, and > some of them can take a long time to complete. We did analyze features > like the query planner's output, our query properties (type, >

Re: [GENERAL] Predicting query runtime

2016-09-12 Thread Vinicius Segalin
2016-09-12 17:01 GMT-03:00 Jeff Janes : > On Mon, Sep 12, 2016 at 7:03 AM, Vinicius Segalin > wrote: > >> Hi everyone, >> >> I'm trying to find a way to predict query runtime (I don't need to be >> extremely precise). I've been reading some papers

Re: [GENERAL] Predicting query runtime

2016-09-12 Thread Vinicius Segalin
2016-09-12 15:16 GMT-03:00 Merlin Moncure : > On Mon, Sep 12, 2016 at 9:03 AM, Vinicius Segalin > wrote: > > Hi everyone, > > > > I'm trying to find a way to predict query runtime (I don't need to be > > extremely precise). I've been reading some papers

Re: [GENERAL] Predicting query runtime

2016-09-12 Thread Istvan Soos
Hi Vinicius, At Heap we have non-trivial complexity in our analytical queries, and some of them can take a long time to complete. We did analyze features like the query planner's output, our query properties (type, parameters, complexity) and tried to automatically identify factors that

Re: [GENERAL] Predicting query runtime

2016-09-12 Thread Jeff Janes
On Mon, Sep 12, 2016 at 7:03 AM, Vinicius Segalin wrote: > Hi everyone, > > I'm trying to find a way to predict query runtime (I don't need to be > extremely precise). I've been reading some papers about it, and people are > using machine learning to do so. For the feature

Re: [GENERAL] Predicting query runtime

2016-09-12 Thread Merlin Moncure
On Mon, Sep 12, 2016 at 9:03 AM, Vinicius Segalin wrote: > Hi everyone, > > I'm trying to find a way to predict query runtime (I don't need to be > extremely precise). I've been reading some papers about it, and people are > using machine learning to do so. For the feature

Re: [GENERAL] Predicting query runtime

2016-09-12 Thread Vinicius Segalin
2016-09-12 12:08 GMT-03:00 Oleksandr Shulgin : > On Mon, Sep 12, 2016 at 4:03 PM, Vinicius Segalin > wrote: > >> Hi everyone, >> >> I'm trying to find a way to predict query runtime (I don't need to be >> extremely precise). I've been reading

Re: [GENERAL] Predicting query runtime

2016-09-12 Thread Oleksandr Shulgin
On Mon, Sep 12, 2016 at 4:03 PM, Vinicius Segalin wrote: > Hi everyone, > > I'm trying to find a way to predict query runtime (I don't need to be > extremely precise). I've been reading some papers about it, and people are > using machine learning to do so. For the feature

[GENERAL] Predicting query runtime

2016-09-12 Thread Vinicius Segalin
Hi everyone, I'm trying to find a way to predict query runtime (I don't need to be extremely precise). I've been reading some papers about it, and people are using machine learning to do so. For the feature vector, they use what the DBMS's query planner provide, such as operators and their cost.