Hello,
just in case someone other sees the same stuff:
A quick look in the source-code, some investigations and a bit of grey
matter revealed that this seems not to be a fatal error.
First we see, that the current request of Task 91 is cancelled:
2006-01-02 11:04:36 6025 11577 COMMUNIC Cancel request for T91
2006-01-02 11:04:36 6036 53000 B*TREE BD500: Next cancelled: 91
This message is generated by the method cbd500_Tree::bd500NextNode which is
used to get the next node of a tree structure. In this case the transaction
was cancelled and therefore this method returns immediatly without returning
the next node. But before returning, it sets 'm_TrError' to 'e_cancelled'.
2006-01-02 11:04:36 6036 ERR 53000 OBJECT BD550FieldQueue:PageLeak: 113
2006-01-02 11:04:36 6036 ERR 53000 OBJECT BD550FieldQueue:PageLeak: 19
2006-01-02 11:04:36 6036 ERR 53000 OBJECT BD550FieldQueue:PageLeak: 54
These messages are caused by the method
cbd550_FieldQueue::~cbd550_FieldQueue which is the destructor of the class
cbd550_FieldQueue. There's a do-while loop in this method to free the nodes
held by the queue. This loop has the condition "e_Ok == m_TrError". Because
cbd500_Tree::bd500NextNode has set 'm_TrError' to 'e_cancelled', the loop
exits prior all nodes are freed.
Afterwards there's a check if all nodes have been released. This wasn't the
case; so these messages with the current page-count are logged:
g01abort (csp3_bd_msg, csp3_n_obj,
"BD550FieldQueue:PageLeak", m_AllocPageCnt);
Because of these facts you should consider to maybe restart your instance
(don't know whether these nodes are freed by a kind of garbage-collector).
But there's no inconsistency in e.g. the devspaces as feared by me.
NOTE: If this happens on your SAP R/3 System you should contact SAP Support
to analyse this problem exactly.
Bye
Chris
>-----Original Message-----
>From: JUNG, Christian
>Sent: Tuesday, January 03, 2006 11:22 AM
>To: MaxDB Mailinglist (E-Mail)
>Subject: knldiag.err: BD550FieldQueue:PageLeak
>
>
>Hello,
>
>we had some problems in a SAP R/3-System running with a MaxDB
>7.5.0.30 on
>Linux. First I thought about some problems with the database
>and therefore
>took a look in the log-files.
>
>I've found these messages which don't make much sense to me in
>'knldiag':
>
>---%<---
>2006-01-02 11:04:36 6025 11577 COMMUNIC Cancel request for T91
>2006-01-02 11:04:36 6036 53000 B*TREE BD500: Next cancelled: 91
>2006-01-02 11:04:36 6036 ERR 53000 OBJECT
>BD550FieldQueue:PageLeak: 113
>2006-01-02 11:04:36 6036 ERR 53000 OBJECT
>BD550FieldQueue:PageLeak: 19
>2006-01-02 11:04:36 6036 ERR 53000 OBJECT
>BD550FieldQueue:PageLeak: 54
>---%<---
>
>I don't think that's the cause of the R/3-problem but I'm
>curious to get an
>explanation of what happened there. The database runs fine, there a no
>problems and a quick search via google didn't reveal anything.
>Do I have to
>worry about these messages?
>
>The exactly build is 'X32/LINUX 7.5.0 Build 030-121-100-791'
>running on SuSE
>SLES 9 with Linux SMP-kernel 2.6.5-7 build 191.
>
>
>Greets & a happy new year
>Chris
>
>phone: +49 6898/10-4987
>fax: +49 6898/10-54987
>http://www.saarstahl.de
>
>--
>MaxDB Discussion Mailing List
>For list archives: http://lists.mysql.com/maxdb
>To unsubscribe:
http://lists.mysql.com/[EMAIL PROTECTED]
--
MaxDB Discussion Mailing List
For list archives: http://lists.mysql.com/maxdb
To unsubscribe: http://lists.mysql.com/[EMAIL PROTECTED]
