| Bug #25389 | Crash on ArrayPool<T>::getPtr | ||
|---|---|---|---|
| Submitted: | 3 Jan 2007 15:04 | Modified: | 21 Sep 2007 22:51 |
| Reporter: | Kris Buytaert (Candidate Quality Contributor) | Email Updates: | |
| Status: | No Feedback | Impact on me: | |
| Category: | MySQL Cluster: Disk Data | Severity: | S3 (Non-critical) |
| Version: | 5.1.14, 5.1.12 | OS: | Linux (Linux) |
| Assigned to: | Assigned Account | CPU Architecture: | Any |
[3 Jan 2007 15:04]
Kris Buytaert
[3 Jan 2007 15:05]
Kris Buytaert
Trace of an ndbd crash
Attachment: traces.log.tar.gz (application/x-gzip, text), 125.18 KiB.
[3 Jan 2007 15:27]
Jonas Oreland
Hi, A number of fixes has been made in 5.1.14 Is it possible for you to test this version ? /Jonas
[3 Jan 2007 16:57]
Valeriy Kravchuk
Please, try to repeat with a newer version, 5.1.14, and inform about the results.
[29 Jan 2007 16:22]
Rob Kinyon
I am able to reproduce this using 5.1.14, but using TEXT columns. As my test setups were in chroot jails, ndb_error_reporter doesn't work. However, the following was the error in the node's errorlog: Time: Monday 29 January 2007 - 15:13:46 Status: Temporary error, restart node Message: Assertion (Internal error, programming error or missing error message, please report a bug) Error: 2301 Error data: ArrayPool<T>::getPtr Error object: ../../../../../storage/ndb/src/kernel/vm/ArrayPool.hpp line: 396 (block: DBLQH) Program: ./ndbd Pid: 28725 Trace: /var/lib/mysql-cluster/ndb_2_trace.log.1 Version: Version 5.1.14 (beta) ***EOM*** This occurs right at startup. If I changed my one TEXT column to VARCHAR(2000), I could not replicate. This also only occurs if there is actual data in the TEXT column. If the table is empty, then there is no problem.
[13 Mar 2007 16:58]
Nathan Babb
I'm seeing similar behavior. I too had data in the table that I was trying to update. Inserts work fine. I'll attach the output from ndb_error_reporter.
The statement that caused the crash: update IndexEntry set text = "Paris Bluegreen" where id=34225
The DDL for the table:
CREATE
TABLE IndexEntry
(
id bigint(20) NOT NULL AUTO_INCREMENT,
text varchar(255) NOT NULL,
deleted bit(1) NOT NULL,
modified datetime,
sortText varchar(255),
PRIMARY KEY USING BTREE (id)
)
ENGINE= NDBCLUSTER DEFAULT CHARSET= utf8 COMMENT= 'number_of_replicas: 2'
[13 Mar 2007 16:59]
Nathan Babb
NDB Error Report
Attachment: ndb_error_report_20070313125749.tar.bz2 (application/octet-stream, text), 136.31 KiB.
[13 Mar 2007 16:59]
Nathan Babb
Also, I am on 5.1.16-beta.
[8 May 2007 8:33]
Jonas Oreland
Hi Kris, Rob & Nathan Do any one of you have this repeatable. I.e a set of schema+data+sql that will reproduce problem with some likelihood /Jonas
[21 Aug 2007 22:51]
Hartmut Holzgraefe
I wasn't able to reproduce this either using
basic setup (2 data nodes, 2 replica):
CREATE LOGFILE GROUP lg_1
ADD UNDOFILE 'undo_1.dat'
INITIAL_SIZE 16M
UNDO_BUFFER_SIZE 2M
ENGINE NDB;
ALTER LOGFILE GROUP lg_1
ADD UNDOFILE 'undo_2.dat'
INITIAL_SIZE 12M
ENGINE NDB;
CREATE
TABLE IndexEntry
(
id bigint(20) NOT NULL AUTO_INCREMENT,
text varchar(255) NOT NULL,
deleted bit(1) NOT NULL,
modified datetime,
sortText varchar(255),
PRIMARY KEY USING BTREE (id)
)
ENGINE= NDBCLUSTER DEFAULT CHARSET= utf8
COMMENT='number_of_replicas: 2';
insert into IndexEntry
set id = 1, text="foobar", deleted=1, modified=now(), sortText="barfoo";
update IndexEntry set text = "Paris Bluegreen" where id=34225;
so we really need some self contained test case
to reproduce this ...
[21 Sep 2007 23:00]
Bugs System
No feedback was provided for this bug for over a month, so it is being suspended automatically. If you are able to provide the information that was originally requested, please do so and change the status of the bug back to "Open".
