Michaelv, thanks for the report.
MySQL or earlier versions of MariaDB are not affected; but I'm afraid if you really need INSERT DELAYED functionality, MySQL is not an option for you, it does not have INSERT DELAYED anymore, it was deprecated in 5.6 and in 5.7 it's treated as a normal INSERT.
The problem was introduced in 10.1.9 with this commit:
commit e3868ee07273b1ea2c495d92891fd7d1ed1eea9e
|
Author: Monty <monty@mariadb.org>
|
Date: Fri Nov 6 13:02:19 2015 +0200
|
|
|
Don't store vcol bitmaps in TABLE if table doesn't have virtual fields.
|
(Makes TABLE a bit smaller)
|
The warning returned upon a successful execution in the output quoted in description is irrelevant – it's just a data type conversion. Below is the same test case, just one value is changed from string to integer to get rid of the warning; it does not affect the outcome, it still fails on current 10.1:
|
|
CREATE TABLE `radacct` ( |
`radacctid` bigint(20) NOT NULL AUTO_INCREMENT, |
`acctsessionid` varchar(64) NOT NULL DEFAULT '', |
`acctuniqueid` varchar(32) NOT NULL DEFAULT '', |
`username` varchar(64) NOT NULL DEFAULT '', |
`groupname` varchar(64) NOT NULL DEFAULT '', |
`realm` varchar(64) DEFAULT '', |
`nasipaddress` varchar(15) NOT NULL DEFAULT '', |
`nasportid` varchar(15) DEFAULT NULL, |
`nasporttype` varchar(32) DEFAULT NULL, |
`acctstarttime` datetime DEFAULT NULL, |
`acctstoptime` datetime DEFAULT NULL, |
`acctsessiontime` int(11) DEFAULT NULL, |
`acctauthentic` varchar(32) DEFAULT NULL, |
`connectinfo_start` varchar(50) DEFAULT NULL, |
`connectinfo_stop` varchar(50) DEFAULT NULL, |
`acctinputoctets` bigint(20) DEFAULT NULL, |
`acctoutputoctets` bigint(20) DEFAULT NULL, |
`calledstationid` varchar(50) NOT NULL DEFAULT '', |
`callingstationid` varchar(50) NOT NULL DEFAULT '', |
`acctterminatecause` varchar(32) NOT NULL DEFAULT '', |
`servicetype` varchar(32) DEFAULT NULL, |
`framedprotocol` varchar(32) DEFAULT NULL, |
`framedipaddress` varchar(15) NOT NULL DEFAULT '', |
`acctstartdelay` int(11) DEFAULT NULL, |
`acctstopdelay` int(11) DEFAULT NULL, |
`acctupdatetime` datetime DEFAULT NULL, |
`deleted` char(1) DEFAULT '', |
`objectclass` varchar(64) DEFAULT 'AAA-SessionInfo', |
PRIMARY KEY (`radacctid`), |
KEY `username` (`username`), |
KEY `framedipaddress` (`framedipaddress`), |
KEY `acctsessionid` (`acctsessionid`), |
KEY `acctsessiontime` (`acctsessiontime`), |
KEY `acctuniqueid` (`acctuniqueid`), |
KEY `acctstarttime` (`acctstarttime`), |
KEY `acctstoptime` (`acctstoptime`), |
KEY `nasipaddress` (`nasipaddress`), |
KEY `acctupdatetime` (`acctupdatetime`), |
KEY `deleted` (`deleted`), |
KEY `objectclass` (`objectclass`) |
) ENGINE=MyISAM AUTO_INCREMENT=7117 DEFAULT CHARSET=latin1; |
|
|
INSERT DELAYED INTO radacct (acctsessionid, acctuniqueid, username, groupname, realm, nasipaddress, nasportid, nasporttype, acctstarttime, acctstoptime, acctsessiontime, acctauthentic, connectinfo_start, connectinfo_stop, acctinputoctets, acctoutputoctets, calledstationid, callingstationid, acctterminatecause, servicetype, framedprotocol, framedipaddress, acctstartdelay, acctstopdelay, acctupdatetime ) VALUES ("aasda1","ascac","ascasc","assfa","reasas","ascac","123","asc",now(),now(),12312,"testa","123asc",10,10,10,"ascasc","teast","ascasc","ascasc","aasca","teast",12312,123123,now()); |
Hi Elena,
Thanks for the fast response. One of the reasons for Insert Delayed was the buffering of insert into the table which
speeds up response time. I wasn't sure I can achieve near the same speed with Innodb/Xtradb.
Hence my testing.
Also I have this weird issue.
I am running a 64bit kernel, running on 8G of ram (VM centos). For some weird reason, when I try to set the innodb_buffer_pool > 2G I get this error on start up.
2016-02-24 16:28:18 4160034560 [Note] /usr/sbin/mysqld (mysqld 10.1.11-MariaDB) starting as process 8897 ...
|
2016-02-24 16:28:18 4160034560 [ERROR] innobase_buffer_pool_size can't be over 4GB on 32-bit systems
|
2016-02-24 16:28:18 4160034560 [ERROR] Plugin 'InnoDB' init function returned error.
|
2016-02-24 16:28:18 4160034560 [ERROR] Plugin 'InnoDB' registration as a STORAGE ENGINE failed.
|
2016-02-24 16:28:18 4160034560 [Note] Plugin 'FEEDBACK' is disabled.
|
2016-02-24 16:28:18 4160034560 [ERROR] Unknown/unsupported storage engine: InnoDB
|
2016-02-24 16:28:18 4160034560 [ERROR] Aborting
|
Kernel being used
Linux server1.testing.com 2.6.18-308.1.1.el5 #1 SMP Wed Mar 7 04:16:51 EST 2012 x86_64 x86_64 x86_64 GNU/Linux
|
$ file /sbin/init
|
/sbin/init: ELF 64-bit LSB executable, AMD x86-64, version 1 (SYSV), for GNU/Linux 2.6.9, dynamically linked (uses shared libs), for GNU/Linux 2.6.9, stripped
|
Not sure why it is reporting this 32bit issue.
I installed.
-rw-r--r-- 1 root root 11246365 Jan 29 12:40 galera-25.3.12-1.rhel5.el5.i386.rpm
|
-rw-r--r-- 1 root root 110818 Feb 14 2015 jemalloc-3.6.0-2.el5.i386.rpm
|
-rw-r--r-- 1 root root 14409006 Jan 29 13:32 MariaDB-10.1.11-centos5-i686-client.rpm
|
-rw-r--r-- 1 root root 51351 Jan 29 13:32 MariaDB-10.1.11-centos5-i686-common.rpm
|
-rw-r--r-- 1 root root 3491380 Jan 29 13:32 MariaDB-10.1.11-centos5-i686-compat.rpm
|
-rw-r--r-- 1 root root 2122913 Jan 29 13:32 MariaDB-10.1.11-centos5-i686-connect-engine.rpm
|
-rw-r--r-- 1 root root 69106584 Jan 29 13:32 MariaDB-10.1.11-centos5-i686-server.rpm
|
-rw-r--r-- 1 root root 1653188 Jan 29 13:32 MariaDB-10.1.11-centos5-i686-shared.rpm
|
In one of your replies you stated.
Fix Version/s 10.1 [ 16100 ]
Assignee Michael Widenius [ monty ]
Does this means it is already fixed? Our its being assigned to Michael Widenius to be fixed?
Regards,
Michael
Opps I mean 4G not > 2G 
Michaelv, you have installed 32-bit packages of MariaDB (i686). 64-bit packages have x86_64 in their name.
In one of your replies you stated.
Fix Version/s 10.1 [ 16100 ]
Assignee Michael Widenius [ monty ]
Does this means it is already fixed? Our its being assigned to Michael Widenius to be fixed?
No, it hasn't been fixed it. When it's fixed, the Resolution value will be changed to Fixed, and Status to Closed.
To monty: My first comment to the issue relates to this bug report, the rest is irrelevant and can be ignored.
The problem was that I had accidently typed + instead of * when initializing a bit array.
Fix pushed into 10.1.