"show innodb status" => Zeos: row buffer width excedeed
Brought to you by:
ansgarbecker
Originally created by: Zimmerma...@gmail.com
What exact steps will reproduce the problem?
1. select windows "Query"
2. show innodb status
3. execute
What was the expected output?
zeos: row buffer width excedeed
What happened instead?
Suggested fix (optional)?
Version used?
HeidiSQL revision:
MySQL Server version:
Operating system:
View and moderate all "tickets Discussion" comments posted by this user
Mark all as spam, and block user from posting to "Tickets"
Originally posted by: a...@anse.de
Not reproducable here. MySQL version? HeidiSQL version?
Labels: Severity-Broken
Status: NeedInfo
View and moderate all "tickets Discussion" comments posted by this user
Mark all as spam, and block user from posting to "Tickets"
Originally posted by: Zimmerma...@gmail.com
Version used? 4
HeidiSQL revision: 2448
MySQL Server version: 5.1
Operating system: Win XP
View and moderate all "tickets Discussion" comments posted by this user
Mark all as spam, and block user from posting to "Tickets"
Originally posted by: a...@anse.de
(No comment was entered for this change.)
Status: Accepted
View and moderate all "tickets Discussion" comments posted by this user
Mark all as spam, and block user from posting to "Tickets"
Originally posted by: a...@anse.de
(No comment was entered for this change.)
Summary: "show innodb status" => Zeos: row buffer width excedeed
View and moderate all "tickets Discussion" comments posted by this user
Mark all as spam, and block user from posting to "Tickets"
Originally posted by: rosenfie...@gmail.com
I think the row buffer is about 64KiB in size, per row.
The driver part of Zeos should take columns which can prospectively contain fields
that are larger than this (actually, some smaller length, to account for the fact
that there are multiple columns) and request to store them as referenced objects
instead.
The mechanism breaks if you select "too many" columns, that is if the "smaller
length" limit mentioned above multiplied by the number of columns exceed the total
per-row size of the row buffer.
Another way this could break is if the server lies about the maximum field size when
sending metadata for a column. Zeos would allocate a buffer and later get a field
larger than what was specified as the maximum for that column.
Either way, I think the problem would be easier to debug if you could obtain a
protocol dump showing the SHOW INNODB STATUS and result. One way to do this is with
HeidiSQL and a .cap file from Wireshark; the login message is challenge-response
encoded so your password will not be divulged, but you may want to filter out the
login message anyway using fx a "frame.number > 123" display filter and saving only
filtered packets. Another way is to run SHOW INNODB STATUS under the official mysql
CLI with the -T switch enabled and copy/paste a complete transcript of that.
NeedInfo: can't reproduce locally, need dump with problematic data.
Status: NeedInfo