Bug #11739 SendBufferMemory set to 294967039 causes core where max = 4294967039
Submitted: 5 Jul 2005 17:31 Modified: 13 Oct 2005 11:16
Reporter: Jonathan Miller Email Updates:
Status: Closed Impact on me:
Category:MySQL Cluster: Cluster (NDB) storage engine Severity:S3 (Non-critical)
Version:4.1,5.0,5.1 OS:Linux (Linux)
Assigned to: Tomas Ulin CPU Architecture:Any

[5 Jul 2005 17:31] Jonathan Miller
Trying Jonas's suggestion of increasing SendBufferMemory in ref to "MySQL Bugs: #11730: Sabre test cause cluster to go down. SendBufferMemory should be dynamic" I increased the buffer size to 524288 and still had the cluster go down due to maxing out the current buffer size. I then increased the value to 1048576 and still had the same issues.

So since the Legal values are between 65536 and 4294967039 I set the buffer size to 294967039. This causes all the ndbd processes to core on startup,

Date/Time: Tuesday 5 July 2005 - 19:17:00
Type of error: error
Message: Internal program error (failed ndbrequire)
Fault ID: 2341
Problem data: BackupInit.cpp
Object of reference: BACKUP (Line: 111) 0x0000000a
ProgramName: /home/ndbdev/jmiller/builds/libexec/ndbd
ProcessID: 10039
TraceFile: /space/run/ndb_3_trace.log.3
Version 5.1.0 (a_drop5p3)

All files will be saved off to ndb08: /space/bug#### where #### = the number of this reports. 

Note there are about 8 - 9 core files for problem debugging stored there.

How to repeat:
Use the following config.ini and start the cluster.
NoOfReplicas: 2
IndexMemory: 100M
DataMemory: 600M
BackupMemory: 64M
MaxNoOfConcurrentScans: 100
DataDir: /space/run
FileSystemPath: /space/run

PortNumber: 14000
ArbitrationRank: 1
DataDir: .
SendBufferMemory: 294967039
Id: 1
HostName: ndb08

Id: 2
HostName: ndb08

Id: 3
HostName: ndb09

Id: 4
HostName: ndb08

Id: 5
HostName: ndb09

Id: 6
HostName: ndb08

Id: 7
HostName: ndb08

Id: 8
HostName: ndb08

Id: 9
HostName: ndb09

Id: 10
HostName: ndb09

Id: 11
HostName: ndb09

Id: 14
HostName: ndb08

Id: 15
HostName: ndb09

Suggested fix:
Not sure, just want to be able to run the Sabre test without the nodes going down due to SendBufferMemory size being too small and nodes coreing because I increase the value,
[20 Jul 2005 6:43] Tomas Ulin
This is most certainly a configuration error.  And most certainly also in 4.1

Setting lower priority.
[4 Oct 2005 7:29] Bugs System
A patch for this bug has been committed. After review, it may
be pushed to the relevant source trees for release in the next
version. You can access the patch from:

[10 Oct 2005 6:08] Tomas Ulin
Bug #13197, Bug #11749, Bug #12786, Bug #11739

New and improved Error messages from 5.0.15.  Error messages not updated in 4.1.
[13 Oct 2005 11:16] Jon Stephens
Thank you for your bug report. This issue has been committed to our
source repository of that product and will be incorporated into the
next release.

If necessary, you can access the source repository and build the latest
available version, including the bugfix, yourself. More information 
about accessing the source trees is available at

Additional info:

Documented in 5.0.15 changelog. Closed.