[Firebird-devel] [FB-Tracker] Created: (CORE-3922) Sweep does not start if one of connects doing select * from ... where ... FOR UPDATE WITH LOCK without COMMIT or ROLLBACK
Sweep does not start if one of connects doing select * from ... where ... FOR
UPDATE WITH LOCK without COMMIT or ROLLBACK
--
Key: CORE-3922
URL: http://tracker.firebirdsql.org/browse/CORE-3922
Project: Firebird Core
Issue Type: Bug
Components: Engine
Reporter: Pavel Zotov
Attachments: trace_no_sweep_starts_when_select_for_update_with_lock.zip
session #1
##
C:\1INSTALL\FIREBIRD\Data>isql localhost:C:\1INSTALL\FIREBIRD\Data\t1.fdb -n
Database: localhost:C:\1INSTALL\FIREBIRD\Data\t1.fdb
SQL> show database;
Database: localhost:C:\1INSTALL\FIREBIRD\Data\t1.fdb
Owner: SYSDBA
PAGE_SIZE 4096
Number of DB pages allocated = 196
Sweep interval = 2
Forced Writes are OFF
Transaction - oldest = 3
Transaction - oldest active = 4
Transaction - oldest snapshot = 4
Transaction - Next = 6
ODS = 11.2
Default Character set: NONE
recreate table tfix1(id int primary key, s01 varchar(1000));
recreate table tfix2(id int primary key, s01 varchar(1000));
recreate table tfix3(id int primary key, s01 varchar(1000));
commit;
insert into tfix1 values(-1,'tfix1');
insert into tfix2 values(-2,'tfix2');
insert into tfix3 values(-2,'tfix2');
commit;
set transaction read committed;
set term ^;
execute block as
declare n int = 20;
declare k int;
begin
while (n>0) do begin
insert into tfix1(id,s01) values(:n, rpad('',1000,'x'));
insert into tfix2(id,s01) values(:n, rpad('',1000,'x'));
insert into tfix3(id,s01) values(:n, rpad('',1000,'x'));
n=n-1;
end
end^
set term ;^
-- here we must wait a few minutes until this EB is finished ...
session #2
##
session #2
C:\1INSTALL\FIREBIRD\Data>isql localhost:C:\1INSTALL\FIREBIRD\Data\t1.fdb -n
Database: localhost:C:\1INSTALL\FIREBIRD\Data\t1.fdb
commit;
set transaction read committed;
set term ^;
execute block as
declare n int = 21000;
declare k int;
begin
while (n>0) do begin
in autonomous transaction do select 1 from rdb$database into :k;
n=n-1;
end
end^
set term ;^
commit;
set list on;
select * from tfix1 where id<0 for update with lock;
set list off;
/*
after 1..2 seconds we will get:
ID -1
S01 tfix1
*/
session #3
###
start trace with enabled log_sweep.
session #1
##
SQL> rollback;
SQL> show database; -- this action finishes with COMMIT
Database: localhost:C:\1INSTALL\FIREBIRD\Data\t1.fdb
Owner: SYSDBA
PAGE_SIZE 4096
Number of DB pages allocated = 11810
Sweep interval = 2
Forced Writes are OFF
Transaction - oldest = 5
Transaction - oldest active = 21009
Transaction - oldest snapshot = 6
Transaction - Next = 21011
ODS = 11.2
Default Character set: NONE
SQL> select 1 from rdb$database;
CONSTANT
1
-- sweep does NOT start here (no info about it in trace window)
session #2
##
SQL> commit;
SQL> set list on;
SQL> select * from tfix1 where id<0 for update with lock; -- sweep starts only
now
ID -1
S01 tfix1
SQL> set list off;
See firebird.log and trace in attach.
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators:
http://tracker.firebirdsql.org/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
--
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
Firebird-Devel mailing list, web interface at
https://lists.sourceforge.net/lists/listinfo/firebird-devel
[Firebird-devel] [FB-Tracker] Created: (CORE-3921) "record disappeared (186), file: vio.cpp line: 408" + CPU 100% (when bugcheckabort=1 and sweep starts at gap ~21000)
"record disappeared (186), file: vio.cpp line: 408" + CPU 100% (when
bugcheckabort=1 and sweep starts at gap ~21000)
-
Key: CORE-3921
URL: http://tracker.firebirdsql.org/browse/CORE-3921
Project: Firebird Core
Issue Type: Bug
Components: Engine
Reporter: Pavel Zotov
Attachments: trace_sweep_vio_error.zip
Intro: create database 't0.fdb'; commit; quit;
session #1
##
C:\1INSTALL\FIREBIRD\Data>isql localhost:C:\1INSTALL\FIREBIRD\Data\t0.fdb -n
Database: localhost:C:\1INSTALL\FIREBIRD\Data\t0.fdb
SQL> recreate table tfix1(id int primary key, s01 varchar(1000));
SQL> recreate table tfix2(id int primary key, s01 varchar(1000));
SQL> commit;
SQL> insert into tfix1 values(-1,'tfix1');
SQL> insert into tfix2 values(-2,'tfix2');
SQL> commit;
SQL>
SQL> set transaction read committed;
SQL> set term ^;
SQL> execute block as
CON> declare n int = 20;
CON> declare k int;
CON> begin
CON> while (n>0) do begin
CON> --insert into ttmp(id,s01) values(:n, rpad('',1000,'x'));
CON> insert into tfix1(id,s01) values(:n, rpad('',1000,'x'));
CON> insert into tfix2(id,s01) values(:n, rpad('',1000,'x'));
CON> n=n-1;
CON> end
CON> end^
SQL> set term ;^
session #2 // start it after execute block in session #1 has finished
##
commit;
set transaction read committed;
set term ^;
execute block as
declare n int = 21000;
declare k int;
begin
while (n>0) do begin
in autonomous transaction do select 1 from rdb$database into :k;
n=n-1;
end
end^
set term ;^
commit;
session #1 // switch in it only after EB in session #2 has finished
##
rollback;
/* at this point we have:
Database "T0.FDB"
Database header page information:
<...>
Page size 4096
ODS version 11.2
Oldest transaction 5
Oldest active 6
Oldest snapshot 6
Next transaction21008
<...>
Database dialect3
Creation date Sep 11, 2012 8:09:28
Attributes force write
Variable header data:
*END*
*/
session #3
##
Start user trace with the following config file:
enabled true
log_connections true
log_transactions true
log_statement_prepare true
log_statement_free true
log_statement_start true
log_statement_finish true
# log_trigger_start true
log_trigger_finish true
log_blr_requests true
log_errors true
log_sweep true # << ENABLE sweep info to be logged
<<<
print_perf true
print_blr true
time_threshold 0
max_sql_length 4000
max_blr_length 1000
max_dyn_length 1000
max_log_size 0
#log_procedure_finish true
session #1
#
update tfix2 set s01='' where id<0; -- стартует sweep, в трейсе по нему
появляется туча сообщений
session #2
#
SQL> select count(*) from tfix1;
COUNT
1
After this point I got standard 'critical error windows' on my Windows-2000
machine (fb_inet_server.exe - The application has requested the Runtime to
terminate it in an unusial way).
Trace window has been stopped (like if I press Ctrl-C in it) and the following
messages appeared in firebird.log:
BALAHA Tue Sep 11 08:25:40 2012
Sweep is started by SWEEPER
Database "C:\1INSTALL\FIREBIRD\DATA\T0.FDB"
OIT 5, OAT 6, OST 6, Next 21009
BALAHA Tue Sep 11 08:25:50 2012
Database: C:\1INSTALL\FIREBIRD\DATA\T0.FDB
internal Firebird consistency check (record disappeared (186), file:
vio.cpp line: 408)
BALAHA (Client) Tue Sep 11 08:26:11 2012
XNET error: Server shutdown detected
No any actions could be done in both the sessions #1 and 2:
SQL> select count(*) from tfix1;
Statement failed, SQLSTATE = 08006
Unable to complete network request to host "localhost".
-Error writing data to the connection.
Trace text has been attached in ZIP.
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators:
http://tracker.firebirdsql.org/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
--
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
Firebird-Devel mailing list, web interface at
https://lists.sourceforge.net/lists/listinfo/firebird-devel
[Firebird-devel] [FB-Tracker] Created: (CORE-3920) Updating Null Column values Will Raise an Exception -605
Updating Null Column values Will Raise an Exception -605 Key: CORE-3920 URL: http://tracker.firebirdsql.org/browse/CORE-3920 Project: Firebird Core Issue Type: Bug Components: Engine Affects Versions: 2.1.3 Environment: Windows, Delphi, IBOConsole, AMDTurion X2 Reporter: Sérgio ENDDATE is date and was created after the table being created and it allows nulls so it filled the colum with nulls. So i want to set ENDDATE then i do : UPDATE MYTABLE SET ENDDATE=CURRENT_DATE WHERE MYID=1 Error -605 value null This wasnt supposed to happen ? It should Set RowID MYID primarykey ENDDATE with th CURRENT_DATE .The problem yet is in the IBOConsole and My Application Components if i change it manualy it accepts it with no problem only Query Problem. I dont undestand wheres the problem. -- This message is automatically generated by JIRA. - If you think it was sent incorrectly contact one of the administrators: http://tracker.firebirdsql.org/secure/Administrators.jspa - For more information on JIRA, see: http://www.atlassian.com/software/jira -- Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ Firebird-Devel mailing list, web interface at https://lists.sourceforge.net/lists/listinfo/firebird-devel
Re: [Firebird-devel] Restore improvements?
Thomas, > I'm currently running a restore of a Firebird 2.5.2 TPC-H scale 1 database > (both, backup and restore are on a consumer SDD) and in that case the index > creation process is mainly CPU bound, basically fully utilizing 1 core of my > hexa-core desktop. Are you testing using IB XE3? Being CPU bound is what I would expect. > Is this something where Firebird could benefit as well? If yes, would this be > possible across all architectures or does this need a shared page cache etc. Yes, this is something that FB could benefit from. Yes, it should be possible across all architectures. No, it should not require a shared page cache. I don't think the cache factors into it. > Another improvement for the index creation process during a restore might > be to allow a temporary bigger page cache, so perhaps a -cache, -buffers or > whatever option might be helpful. Pretty much what e.g isql already allows at > connection level etc. I don't see how a larger cache or more buffers would help the index creation process. The ideal process would have a single pass of a table collect data from columns in all indexes, into a temporary file (or multiple), then pass that data to appropriate sorting/index create routines. In which case, the number of pages not going to make a significant difference because each page would be read only once. The cache is beneficial when multiple page accesses are involved. Sean -- Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ Firebird-Devel mailing list, web interface at https://lists.sourceforge.net/lists/listinfo/firebird-devel
[Firebird-devel] Restore improvements?
Hello, InterBase XE3 got something which allows to create indexes during restore in parallel. http://docwiki.embarcadero.com/RADStudio/en/InterBase_XE3 I'm currently running a restore of a Firebird 2.5.2 TPC-H scale 1 database (both, backup and restore are on a consumer SDD) and in that case the index creation process is mainly CPU bound, basically fully utilizing 1 core of my hexa-core desktop. Is this something where Firebird could benefit as well? If yes, would this be possible across all architectures or does this need a shared page cache etc. Another improvement for the index creation process during a restore might be to allow a temporary bigger page cache, so perhaps a -cache, -buffers or whatever option might be helpful. Pretty much what e.g isql already allows at connection level etc. Thoughts? Thanks, Thomas -- Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ Firebird-Devel mailing list, web interface at https://lists.sourceforge.net/lists/listinfo/firebird-devel
