summaryrefslogtreecommitdiff
path: root/KNOWN_BUGS.txt
diff options
context:
space:
mode:
authorMichael Widenius <monty@mysql.com>2008-06-28 18:09:03 +0300
committerMichael Widenius <monty@mysql.com>2008-06-28 18:09:03 +0300
commitf27efe69f50a2c60104fca3a7937c37ce48da592 (patch)
treeac34b63b5e9c9df801f2ed0274bb8eca19863e6d /KNOWN_BUGS.txt
parentc580583756a6c56d00245d63663ce38e28a7023c (diff)
downloadmariadb-git-f27efe69f50a2c60104fca3a7937c37ce48da592.tar.gz
Disable versioning (concurrent writes) if more than one unique key
The reason for this is that if we change one unique key and then get a failure on the second, we may not be able to rename the first one back before someone else writes the same key value. In Maria 2.0, when we keep deleted key values in the tree, this will not be a problem anymore Fixed typedisable concurrent insert/select for SQLCOM_LOAD as there are problems with concurrent threads during index recreation KNOWN_BUGS.txt: More comments storage/maria/ha_maria.cc: Fixed typo (REPLACE -> INSERT) Also disable concurrent insert/select for SQLCOM_LOAD as there are problems with concurrent threads during index recreation storage/maria/ma_open.c: Disable versioning (concurrent writes) if more than one unique key
Diffstat (limited to 'KNOWN_BUGS.txt')
-rw-r--r--KNOWN_BUGS.txt15
1 files changed, 13 insertions, 2 deletions
diff --git a/KNOWN_BUGS.txt b/KNOWN_BUGS.txt
index 980dd5f0da9..a69384910e2 100644
--- a/KNOWN_BUGS.txt
+++ b/KNOWN_BUGS.txt
@@ -29,6 +29,11 @@ Known bugs that we are working on and will be fixed shortly
the log handler doesn't start up after restart.
Most of this should now be fixed...
+- INSERT on a duplicate key against a key inserted by another connection
+ that has not yet ended will give a duplicate key error instead of
+ waiting for the other statement to end.
+
+
Known bugs that are planned to be fixed before Gamma/RC
=======================================================
@@ -56,14 +61,20 @@ this on a table:
wrong results if someone else is doing writes on the table during repair
or someone is doing selects during the repair index phase.
-INSERT ... SELECT and REPLACE ... SELECT are blocking inserts and
-SELECT for the table.
+INSERT ... SELECT, REPLACE ... SELECT and LOAD DATA are blocking
+inserts and SELECT for the table. They should only have to do this if
+the destination is empty (as then we are using fast index rebuild).
Missing features that is planned to fix before Beta
===================================================
+None
+
Features planned for future releases
====================================
+Most notable is full transaction support and multiple reader/writers
+in Maria 2.0
+
http://forge.mysql.com/worklog/
(you can enter "maria" in the "quick search" field there).