Full_Name: Lars Uffmann Version: 1.2.6 OS: linux 2.2.10 URL: ftp://ftp.openldap.org/incoming/ Submission from: (NULL) (195.71.129.112) In some situations it is possible that slapd will send LDAP_SUCCESS in response to search requests even though an internal (ldbm) backend error occured. The client will recive an empty search result with code LDAP_SUCCESS and has to assume the entry(ies) in question doesn't exist. ldbm backend build with version 2.7.5 (04/18/99) of Sleepycat Software's Berkeley DB. I am not able to reproduce this problem. My specific problem occured after weeks of continous service.
At 10:42 AM 9/29/99 GMT, lars.uffmann@mediaways.net wrote: >Full_Name: Lars Uffmann >Version: 1.2.6 >OS: linux 2.2.10 >URL: ftp://ftp.openldap.org/incoming/ >Submission from: (NULL) (195.71.129.112) > > >In some situations it is possible that slapd will send LDAP_SUCCESS in response >to search requests even though an internal (ldbm) backend error occured. >The client will recive an empty search result with code LDAP_SUCCESS and has to >assume the entry(ies) in question doesn't exist. > >ldbm backend build with version 2.7.5 (04/18/99) of Sleepycat Software's >Berkeley DB. > >I am not able to reproduce this problem. My specific problem occured after >weeks of continous service. The problem is that the some back-ldbm IDL routines have no mechanism to distingish no index from an error. This will require a bit of reworking of the internal interfaces, but should be done for both 2.0 and 1.2.x. I'll add it to the TODO list. Kurt
"Kurt D. Zeilenga" wrote: > > At 10:42 AM 9/29/99 GMT, lars.uffmann@mediaways.net wrote: > >Full_Name: Lars Uffmann > >Version: 1.2.6 > >OS: linux 2.2.10 > >URL: ftp://ftp.openldap.org/incoming/ > >Submission from: (NULL) (195.71.129.112) > > > > > >In some situations it is possible that slapd will send LDAP_SUCCESS in response > >to search requests even though an internal (ldbm) backend error occured. > >The client will recive an empty search result with code LDAP_SUCCESS and has to > >assume the entry(ies) in question doesn't exist. > > > >ldbm backend build with version 2.7.5 (04/18/99) of Sleepycat Software's > >Berkeley DB. > > > >I am not able to reproduce this problem. My specific problem occured after > >weeks of continous service. > > The problem is that the some back-ldbm IDL routines > have no mechanism to distingish no index from an error. > This will require a bit of reworking of the internal > interfaces, but should be done for both 2.0 and 1.2.x. > I'll add it to the TODO list. Can you give me some hints? I would like to work on this. -- Lars Uffmann, <lars.uffmann@mediaways.net>, fon: +49 5241 80 40330
moved from Incoming to Software Bugs
changed notes changed state Open to Suspended
changed state Suspended to Closed
Concur. Requires significant changes