dimitr
e5a70baba9
Backported CORE-2591: High mutex wait ratio and degraded performance after some time of performing normally.
...
The allocation granularity is changed from 4 to 8 bytes.
2009-08-13 14:57:54 +00:00
alexpeshkoff
3ab6a3daa5
Backported Bill's patch for HPUX
2009-08-13 11:19:24 +00:00
alexpeshkoff
6e6ab6b30c
Correction to my fix for CORE-2572
2009-08-13 08:38:09 +00:00
alexpeshkoff
17a218f357
Backported fix for CORE-2572: Locks of type LCK_page_space processed incorrectly
2009-07-24 13:08:14 +00:00
alexpeshkoff
607719c547
Updated startup script from Philippe Makowski
2009-07-21 11:16:57 +00:00
robocop
e5ec513448
Misc.
2009-07-21 09:45:30 +00:00
dimitr
d5e6d69e99
Update.
2009-07-20 14:23:10 +00:00
dimitr
24a4e6f5cd
Update.
2009-07-20 14:13:48 +00:00
dimitr
a095a41dec
Fixed two issues for the case of non-authorized op_connect_request:
...
1) Missing error reporting which makes the client to stall forever
2) CORE-2563: Possible DoS attack using the malformed packet sent into the connection port
2009-07-20 14:06:17 +00:00
alexpeshkoff
e33513173c
Fixed CORE-2564: alignment error on risc machines
2009-07-20 13:43:42 +00:00
alexpeshkoff
77aa52e57a
Fixed NETBSD build
2009-07-20 10:30:19 +00:00
asfernandes
3fe020495c
Misc
2009-07-14 18:56:07 +00:00
dimitr
073debab15
Update.
2009-07-10 08:18:37 +00:00
alexpeshkoff
43584fa0fc
Minimal fix for CORE-2550: Bus error when working with DB_KEY on bigendian machines
2009-07-08 10:18:30 +00:00
dimitr
9782c2cdc5
Correction and trivial cleanup.
2009-07-06 09:18:24 +00:00
dimitr
3a2641724f
Update.
2009-07-06 07:46:24 +00:00
dimitr
5b9a130dc7
Update.
2009-06-29 04:24:46 +00:00
hvlad
9ac27abdca
Fixed bug CORE-2532 : Incorrect size of volumes in multi volume database
2009-06-28 21:11:18 +00:00
dimitr
1f59fa06ef
Fixed CORE-2531: The famous "cannot transliterate" error may be thrown when selecting data from the monitoring tables.
...
This fix doesn't address the non-ASCII user and role names though. Before improving the patch, I need an asnwer whether it's possible in v2.1.
2009-06-28 09:15:12 +00:00
dimitr
670d40bc8e
The trivial fix to get rid of the found regression in v2.1.3 RC1. A complete fix will follow over the weekend.
...
I commit them in two different parts as I dunno what revision is going to be tagged for v2.1.3.
2009-06-26 21:16:25 +00:00
hvlad
1b1b603eb1
Backport fix for bug CORE-2519 : GSTAT output is incorrect for tables with more than 2 billion records
2009-06-21 16:42:12 +00:00
hvlad
f6e2152164
Fixed bug CORE-2507 : Intermittent CreateFile failure
2009-06-20 20:03:02 +00:00
asfernandes
bf23ab7095
Revert change and backport relevant thing from HEAD
2009-06-19 14:42:54 +00:00
paulbeach
74bcaecd78
O_DIRECT and MacOSX problem (lack of) - better ifdefs (I hope)
2009-06-19 14:19:31 +00:00
paulbeach
173b1bfd1a
O_DIRECT and MacOSX problem (lack of)
2009-06-19 13:03:59 +00:00
paulbeach
989b9a9c98
No O_Direct on MacOSX or Solaris. Revert to 2.1.2
2009-06-19 12:36:37 +00:00
paulbeach
cea7e57881
Small mod to installation for Solaris - missing README
2009-06-18 09:21:55 +00:00
paulbeach
8098976e89
Forgotten to port from previous. This ifdef must be here for Solaris (sparc) otherwise we get ld fatal: relocation error R_SPARC_GOT22, relocation illegal for TLS symbol in jrd.cpp
2009-06-16 14:39:57 +00:00
dimitr
0e95a58377
This change allows the valid backup state to be reported via MON$DATABASE. While the nbak_state_unknown state still can be dumped into the shared memory by an AST, it never goes into the monitoring tables due to the special handling for MON$DATABASE.
2009-06-16 06:18:40 +00:00
dimitr
996e07251a
Update.
2009-06-14 15:54:00 +00:00
dimitr
354f5b7cd5
Totally misc.
2009-06-14 04:19:22 +00:00
robocop
585eb2002d
Misc, assertions and const.
2009-06-12 06:21:27 +00:00
dimitr
984da6e742
Misc.
2009-06-11 14:46:12 +00:00
paul_reeves
8757b28281
rollback incorrect change
2009-06-11 12:45:56 +00:00
dimitr
061f3dfbf6
Locking something in AST turns out to be a bad idea. Let's use a "dirty read" approach and honestly return the "unknown" state for a not initialized backup manager (quite rare situation in practice).
2009-06-11 12:16:55 +00:00
dimitr
446534ae8e
Undo the changes no longer necessary.
2009-06-11 11:13:24 +00:00
dimitr
22011cefe5
The third incarnation of the monitoring internals. I'm not happy to commit such big changes into the point release, but this is the only solution for the AST signal safety issues in POSIX (see CORE-2483).
...
The change consists of a few parts:
- get rid of ClumpletReader/Writer classed and the related pool dependencies
- replace a single huge clumplet for dbb with a set of smaller chunks
- work with the dump at the record/field logical level
- use the universal mover instead of manual tweaking with descriptors
- automagical NULL handling instead of the hardcoded nullability rules
So far the testing hasn't revealed any functional or performance regressions.
2009-06-11 11:12:39 +00:00
paul_reeves
173f426561
Update hardcoded version strings
2009-06-10 09:14:52 +00:00
robocop
024d840cf4
Backport improvement CORE-1171: make floating point's exponential format the same across platforms.
2009-06-09 09:59:19 +00:00
alexpeshkoff
c30a601fa5
nbackup.cpp
2009-06-08 13:27:34 +00:00
alexpeshkoff
9861f79f05
Fixed use of monitoring tables in on disconnect trigger, make pool for DatabaseSnapshot independent
2009-06-08 11:47:35 +00:00
alexpeshkoff
de76a6c806
Backported changes extending range of usable auto* tools
2009-06-05 09:50:33 +00:00
dimitr
79db11d0f8
Update.
2009-06-05 09:15:36 +00:00
hvlad
268ee510c5
Update WhatsNew
2009-06-05 08:34:00 +00:00
hvlad
6db905fd86
Backport fix for bug CORE-2348 : More problems with transaction numbers overflowing 32-bit signed integer and corrupting database
2009-06-05 08:25:37 +00:00
hvlad
915429e5fe
Backport sub-task CORE-1647 : Close external table's file even if it used by some cached (inactive) requests.
...
It also fixed bug CORE-2475 : External table data not visible to other sessions in Classic.
2009-06-05 07:25:03 +00:00
hvlad
b1e58c9bbf
Backport fix for bug CORE-2416 : AV preparing a query with aggregate over derived table
2009-06-05 07:08:20 +00:00
dimitr
395abd5842
Update.
2009-06-04 08:58:37 +00:00
dimitr
49ee25df56
Fixed CORE-2483: Database permanent pool may get corrupted when working with monitoring tables.
2009-06-04 08:46:31 +00:00
robocop
ad9cf6e4be
Misc.
2009-05-30 14:23:54 +00:00
