[Date Prev][Date Next] [Chronological] [Thread] [Top]

Re: OpenLDAP 1.0 and Solaris 2.5.1

On Mon, 31 Aug 1998, Kurt D. Zeilenga wrote:

> The attached patch for this bug has commited to -stable and -devel.

Thanks Kurt, that's definitely fixed the first problem.

> >Problem #2 - there's plenty of room on the disk (on all of them!);  OpenLDAP
> >has been compiled with only LDBM backend support (ndbm), nothing else.  Any
> >ideas, or pointers for where to start looking in the code for a solution?
> My guess is that was caused by the test directory not be properly cleaned.

Don't think so - the directory started empty, and after running the tests just

  % ls -Flas test-db
  total 4
     2 drwx------   2 david    staff        512 Sep  1 21:57 ./
     2 drwx------   6 david    staff        512 Sep  1 13:58 ../
     0 -rw-------   1 david    staff          0 Sep  1 21:57 id2entry.ndbm.dir
     0 -rw-------   1 david    staff          0 Sep  1 21:57 id2entry.ndbm.pag

Exactly the same behaviour after your patch for the cleanup has been applied
and OpenLDAP rebuilt.  I added "-d 1" to the test001-ldif2ldbm script (when it
calls ldif2ldbm) and get:

  Initiating LDAP tests...
  >>>>> Executing all LDAP tests...
  >>>>> Starting ./scripts/test001-ldif2ldbm ...
  Cleaning up in ./test-db...
  Running ldif2ldbm to build slapd database...
  => ldbm_cache_open( "./test-db/id2entry.ndbm", 770, 600 )
  <= ldbm_cache_open (opened 0)
  id2entry ldbm_store: No space left on device
  stopping: child exited with status 1
  ldif2ldbm failed!
  >>>>> ./scripts/test001-ldif2ldbm failed (exit 1)

Running with "-d 65535" (LDAP_DEBUG_ANY) didn't really reveal anything new in
the critical area (around the ldbm_store line).  Is that zero ("opened 0") an
index into an array of FDs, an "okay" status value or an indication that the
open itself failed?


David J. N. Begley  <d.begley@nepean.uws.edu.au>
Network Analyst, Communications Unit
University of Western Sydney, Nepean