description: It will be very useful to have ability for quick find in a trace log with huge size some HASH value of statement that was executed.
Consider following sketch:

EXECUTE_STATEMENT_FINISH -------------------------------- [ 1 ]
select <...>
from test1 a
left join test2 b on ...
left join test3 c on ...
...
where <expr_for_a,b,c>
< POOR EXECUTION AND TABLE STATISTICS HERE >

EXECUTE_STATEMENT_FINISH
<... some_other_complex_query_2 ...>

EXECUTE_STATEMENT_FINISH
<... yet_another_complex_query_3...>

EXECUTE_STATEMENT_FINISH -------------------------------- [ 2 ]
select <...>
from test1 a
left join test2 b on ...
left join test3 c on ...
...
where <expr_for_a,b,c>
< POOR EXECUTION AND TABLE STATISTICS HERE >

When someone will read text below [ 2 ] , he can recall that PROBABLY the same query already was already registered before, in [1 ].
But how one may easy and fast to compare such text blocks ? They often take about 50-70 lines of code, and - moreover - this code can be machine-generated.
Of course we can make copy-paste these queries in two diff. files and than compare them. But this is completely inconvenient.

So, please consider ability to add some king of hash value for each executed statement, like this:

2018-02-01T12:20:23.4500 ( <HASH_STRING_FOR_STATEMENT_HERE> ) EXECUTE_STATEMENT_FINISH
prodbase (ATT_29570445, JOHN:MANAGER, UTF8, TCPv4:172.16.190.9)
c:\temp\my_app.exe:4724
(TRA_738419170, READ_COMMITTED | REC_VERSION | WAIT | READ_ONLY)

Statement 108112340:
-------------------------------------------------------------------------------
select count(*) from ...

/*
CURRENTLY within parenthesis there is always the same value (and i don't understand what does it mean). Can we replace it with more useful one ?
*/

PS.
And if it is possible, may be the same hash value can be added for mon$statements.mon$sql_text ?

=>

It will be very useful to have ability for quick find in a trace log with huge size some HASH value of statement that was executed.
Consider following sketch:

EXECUTE_STATEMENT_FINISH -------------------------------- [ 1 ]
select <...>
from test1 a
left join test2 b on ...
left join test3 c on ...
...
where <expr_for_a,b,c>
< POOR EXECUTION AND TABLE STATISTICS HERE >

EXECUTE_STATEMENT_FINISH
<... some_other_complex_query_2 ...>

EXECUTE_STATEMENT_FINISH
<... yet_another_complex_query_3...>

EXECUTE_STATEMENT_FINISH -------------------------------- [ 2 ]
select <...>
from test1 a
left join test2 b on ...
left join test3 c on ...
...
where <expr_for_a,b,c>
< POOR EXECUTION AND TABLE STATISTICS HERE >

When someone will read text below [ 2 ] , he can recall that PROBABLY the same query was already registered before, in [1 ].
But how one may easy and fast to compare such text blocks ? They often take about 50-70 lines of code, and - moreover - this code can be machine-generated.
Of course we can make copy-paste these queries in two diff. files and than compare them. But this is completely inconvenient.

So, please consider ability to add some king of hash value for each executed statement, like this:

2018-02-01T12:20:23.4500 ( <HASH_STRING_FOR_STATEMENT_HERE> ) EXECUTE_STATEMENT_FINISH
prodbase (ATT_29570445, JOHN:MANAGER, UTF8, TCPv4:172.16.190.9)
c:\temp\my_app.exe:4724
(TRA_738419170, READ_COMMITTED | REC_VERSION | WAIT | READ_ONLY)

Statement 108112340:
-------------------------------------------------------------------------------
select count(*) from ...

/*
CURRENTLY within parenthesis there is always the same value (and i don't understand what does it mean). Can we replace it with more useful one ?
*/

PS.
And if it is possible, may be the same hash value can be added for mon$statements.mon$sql_text ?

assignee: Vlad Khorsun [ hvlad ]