Ciao Alessandro,
> regarding INFN-PERUGIA, deleting the ldif files did not changed the situation
>
> They (still) have an lcg-CE sl3, and I've also noticed that on 2135
> port the branch SubCluster is published, look for example:
>
> ldapsearch -x -H ldap://gridce.pg.infn.it:2135 -b
> "mds-vo-name=local,o=grid" objectClass=GlueSubCluster
Indeed, but look at the output of the that command:
------------------------------------------------------------------------------
[...]
GlueSubClusterPhysicalCPUs: GlueSubClusterLogicalCPUs: GlueSubClusterTmpDir: /
tmp
[...]
------------------------------------------------------------------------------
The BDII rejects that record because it violates the schema.
The site should fix those CPU entries and also undo Friday's change:
------------------------------------------------------------------------------
$ uberftp gridce.pg.infn.it 'cat /opt/bdii/etc/bdii-update.conf'
220 gridce.pg.infn.it GridFTP Server 1.12 GSSAPI type Globus/GSI wu-2.6.2
(gcc32dbg, 1062606889-42) ready.
230 User ops017 logged in.
CE ldap://gridce.pg.infn.it:2170/mds-vo-name=resource,o=grid
SE ldap://gridse.pg.infn.it:2135/mds-vo-name=local,o=grid
SE3 ldap://gridse3.pg.infn.it:2135/mds-vo-name=local,o=grid
------------------------------------------------------------------------------
The line for the (gLite 3.0) CE should be:
------------------------------------------------------------------------------
CE ldap://gridce.pg.infn.it:2135/mds-vo-name=local,o=grid
------------------------------------------------------------------------------
For INFN-NAPOLI-ARGO the evidence is in /opt/bdii/var/tmp/stderr.log:
------------------------------------------------------------------------------
str2entry: invalid value for attribute GlueSubClusterPhysicalCPUs
(syntax 1.3.6.1.4.1.1466.115.121.1.27)
------------------------------------------------------------------------------
They have:
------------------------------------------------------------------------------
gluesubclusterphysicalcpus: 0
------------------------------------------------------------------------------
Apparently a SubCluster must have at least 1 CPU... :-)
|