Hi everyone,
Below is some output that I got from upgrading one of my pool/door nodes
using YAIM. There are a couple of things to note:
* There are a lot of references to tomcat in the output. Is tomcat
required for the pool/door node? If so, what for?
* The configure script added a line to my /etc/fstab, such that I had:
srm.epcc.ed.ac.uk:/fs /pnfs/fs nfs
hard,intr,ro,bg,noac,auto 0 0
srm.epcc.ed.ac.uk:/pnfsdoors /pnfs/epcc.ed.ac.uk nfs
hard,intr,ro,bg,noac,auto 0 0
and also removed the link from /pnfs/epcc.ed.ac.uk -> fs/usr meaning that
the pnfs filesytem was no longer available under /pnfs/epcc.ed.ac.uk . I
decided to keep with the second line in /etc/fstab above (since it has
just been added) but this required that I make modifications to the
corresponding /pnfs/fs/admin/exports/<IP-address> file.
* Since I have only upgraded the pool/door software and not the head node,
the gridftp door will not work as it tries to use gPlazma (basically VOMs)
for authorisation. Since dCache 1.7.0 has not yet been installed on the
head node then the relevant gPlazma domain cannot be found.
Here's the error from trying an srmcp:
error: the server sent an error response: 530 530 Authorization Service
failed:
diskCacheV111.services.authorization.AuthorizationServiceException:
authRequestID 497826706No Route to cell for packet
{uoid=<1161075609201:201>;path=[>gPlazma@local];msg=Tunnel cell
>gPlazma@local< not found at >dCacheDomain<}
Matt: this means that if you want to upgrade your door nodes to help with
the CLOSE_WAIT problem then you will also have to upgrade the head node.
I'll let you know if I find anything else.
Cheers,
Greig
<configure_node output>
WARNING: Changed the file /etc/fstab
mkdir: cannot create directory `/pnfs/epcc.ed.ac.uk': File exists
WARNING: Please wait mounting /pnfs/epcc.ed.ac.uk
mount: backgrounding "srm.epcc.ed.ac.uk:/pnfsdoors"
WARNING: Finished mounting /pnfs/epcc.ed.ac.uk
stop config_pnfs_install_mount
end config_sedcache_pnfs
start config_dcache_config_dcache_scripts
Checking MasterSetup ./config/dCacheSetup O.k.
Sanning dCache batch files
Processing adminDoor
Processing dCache
Processing dir
Processing door
Processing gPlazma
Processing gridftpdoor
Processing gsidcapdoor
Processing httpd
Processing infoProvider
Processing lm
Processing pnfs
Processing pool
Processing replica
Processing srm
Processing statistics
Processing utility
mount: mount to NFS server 'srm.epcc.ed.ac.uk' failed: RPC Error:
Success(retrying).
Processing xrootdDoor
Checking Users database .... Ok
Checking Security .... Failed
cd ../config
ssh-keygen -b 768 -f ./server_key -N ""
ssh-keygen -b 1024 -f ./host_key -N ""
Checking JVM ........ Ok
Checking Cells ...... Ok
dCacheVersion ....... Version production-1-7-0
AXIS_PATH=/opt/d-cache/libexec/axis-1_3
TOMCAT_PATH=/opt/d-cache/libexec/jakarta-tomcat-4.1.31
installing tomcat and axis ...Done
rm
/opt/d-cache/libexec/jakarta-tomcat-4.1.31/webapps/srm/WEB-INF/lib/commons-logging*.jar
copying jars ...Done
modifying system CLASSPATH in
/opt/d-cache/libexec/jakarta-tomcat-4.1.31/bin/catalina.sh ...calling sed
CATALINA_CP_ADD=\/opt\/d-cache\/libexec\/jakarta-tomcat-4.1.31\/server\/lib\/cells-protocols.jar:\/opt\/d-cache\/libexec\/jakarta-tomcat-4.1.31\/common\/lib\/cog-url.jar
Done
Starting up tomcat ...Using CATALINA_BASE:
/opt/d-cache/libexec/jakarta-tomcat-4.1.31
Using CATALINA_HOME: /opt/d-cache/libexec/jakarta-tomcat-4.1.31
Using CATALINA_TMPDIR: /opt/d-cache/libexec/jakarta-tomcat-4.1.31/temp
Using JAVA_HOME: /usr/java/j2sdk1.4.2_12
Done
deployng srm v2 application using axis AdminClient ...
- Unable to find required classes (javax.activation.DataHandler and
javax.mail.internet.MimeMultipart). Attachment support is disabled.
Processing file /opt/d-cache/etc/srmv1-deploy.wsdd
<Admin>Done processing</Admin>
- Unable to find required classes (javax.activation.DataHandler and
javax.mail.internet.MimeMultipart). Attachment support is disabled.
Processing file /opt/d-cache/etc/srmv2.2-deploy.wsdd
<Admin>Done processing</Admin>
Done
creating config files and adding configuration info into
/opt/d-cache/libexec/jakarta-tomcat-4.1.31/webapps/srm/WEB-INF/web.xml
...Done
enabling GSI HTTP in tomcat by modifying
/opt/d-cache/libexec/jakarta-tomcat-4.1.31/conf/server.xml
...tmp2:/tmp/tmp2Oct170933116107401110
turning off sending of Multi Refs in
/opt/d-cache/libexec/jakarta-tomcat-4.1.31/webapps/srm/WEB-INF/server-config.wsdd
Done
shutdown Tomcat
Using CATALINA_BASE: /opt/d-cache/libexec/jakarta-tomcat-4.1.31
Using CATALINA_HOME: /opt/d-cache/libexec/jakarta-tomcat-4.1.31
Using CATALINA_TMPDIR: /opt/d-cache/libexec/jakarta-tomcat-4.1.31/temp
Using JAVA_HOME: /usr/java/j2sdk1.4.2_12
installing config for startup/shutdown script
Installation complete
please use /opt/d-cache/bin/dcache-srm start|stop|restart to startup,
shutdown or restart srm server
[INFO] Checking if /pnfs/epcc.ed.ac.uk mounted to the right export. ...
[INFO] Trying to remove symbolic link /pnfs/epcc.ed.ac.uk :
[INFO] 'rm -f /pnfs/epcc.ed.ac.uk' went fine.
[INFO] Creating pnfs mount point (/pnfs/epcc.ed.ac.uk)
[INFO] Will be mounted to srm.epcc.ed.ac.uk:/pnfsdoors by dcache-core
start-up script.
[ERROR] Link /pnfs/ftpBase --> /pnfs/epcc.ed.ac.uk cannot be created.
Needed by the GridFTP door.
Please move /pnfs/ftpBase and run me again. Exiting.
ERROR executing /opt/d-cache/install/install.sh failed (exit code 1) --
abort
[root@pool2 etc]# mount: srm.epcc.ed.ac.uk:/pnfsdoors failed, reason given
by server: Operation not permitted
On Tue, 17 Oct 2006 [log in to unmask] wrote:
> On Mon, 16 Oct 2006 17:53:13 +0100 (BST)
> Greig A Cowan <[log in to unmask]> wrote:
>
> >
> > Hi Owen,
> >
> > That's good to hear that the new version has been released.
> >
> > Is this the correct changelog to look at?
> >
> > http://www.dcache.org/manuals/Book/rf-changelog-1665-1681.shtml
> >
> > I think this page has been online for a while; are there any additional
> > changes/bug fixes that made it into the release? What about the
> > non-blocking I/O that should fix the CLOSE_WAIT problem?
>
> It will fix it to a considerable degree, we found it occures to some degree in an old prerelease but no where near as much as before, we belive it shoudl be fixed now but only time in production will tell.
>
>
> > Presumably you have tested the 1.6.6-5 -> 1.7.0-16 upgrade path. I ask
> > because the link:
> >
> > http://www.dcache.org/manuals/yaim-dcache-howto.shtml
> >
> > states "For now this is only meant for installations from scratch". Were
> > there any other problems during upgrade, for example, the integration
> > of VOMs?
> >
> > Cheers,
> > Greig
>
> Yes its straight forward and waiting for me to finish the instructions, Yaim users see what I wrote before as this is the prefured solution. Bellow is the cut n paste from my README file in the upgrade RPM
>
> REgards
>
> Owen
>
> The procedure for upgrading dCache from 1.6 to 1.7
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>
> Background
> ~~~~~~~~~~
>
> dCache has changed considerably between versions 1.6 and 1.7. Particularly with
> respect to the SRM interface. Please see the change log in the dCache
> documentation at dCache.org for details. These changes have inevitably led to
> many additions to the dCache configuration. This RPM contains a simple script to
> enable sites to easily upgrade thier version of dCache with the minimum effort.
>
> Fortunately the databases of dCache have largely remained unchanged, and so pnfs
> should can be upgraded with little to no effort.
>
> Recipe to upgrade
> ~~~~~~~~~~~~~~~~~
>
> The upgrade must run in three stages, preupgrade, then moving to more modern
> RPM's and then a post upgrade stage.
>
> Preupgrade recipe
> ~~~~~~~~~~~~~~~~~
>
> Run the following commands
>
> /etc/init.d/dcache-pool stop
> /etc/init.d/dcache-core stop
> /etc/init.d/pnfs stop
> dropdb -U srmdcache billing
> dropdb -U srmdcache dcache
> dropdb -U srmdcache replicas
> createdb -U srmdcache billing
> createdb -U srmdcache dcache
> createdb -U srmdcache replicas
> psql -U srmdcache replicas -f /opt/d-cache/etc/psql_install_replicas.sql
>
>
> Rpm Upgrade
> ~~~~~~~~~~~
>
> Now upgrade the RPM's to the 1.7 release of dCache. This may be done via rpm.
>
>
> Postupgrade recipe
> ~~~~~~~~~~~~~~~~~~
>
> The system should now have the 1.7 release of dCache installed. We now need to
> run the configuration upgrade script and start dCache.
>
> sh /opt/d-cache/install/dCacheUpgrade_1_6-1_7.sh
> /etc/init.d/pnfs start
> /opt/d-cache/install/install.sh
> /etc/init.d/dcache-core start
> /etc/init.d/dcache-pool start
>
> Conclusion
> ~~~~~~~~~~
>
> We should now have a fully upgraded dCache. As with previous releases wait a
> 5-10 minutes before you send in any bug reports it may be initialising.
>
>
> >
> > On Mon, 16 Oct 2006 [log in to unmask] wrote:
> >
> > > We have the new stable version of D-cache available on the D-cache web
> > > site,
> > >
> > > Installation instructions are here
> > >
> > > http://www.dcache.org/manuals/yaim-dcache-howto.shtml
> > >
> > > For a YAIM based site you should only need to change the site-info.def
> > > to
> > >
> > > RESET_DCACHE_CONFIGURATION=yes
> > >
> > > While the other settings should be set as
> > >
> > > RESET_DCACHE_PNFS=no
> > > RESET_DCACHE_RDBMS=no
> > >
> > > if you want to keep your data and I guess you do want to keep your data.
> > >
> > > With regard to the apt repository we do not yet have it in stable, I am
> > > giving you all 24 hours notice to check your setups and then I shall
> > > migrate D-cache 1.7.0-16 to stable.
> > >
> > > Non yaim sites should wait for a short time as I finish documenting the
> > > upgrade procedure but its tested and working. I expect this RPM will be
> > > released with instructions in the README at the same time as I release
> > > D-cache 1.7.0-16 to stable in the apt repository.
> > >
> > > D-cache 1.7.0-16 is currently available to YAIM users in the testing
> > > section of the apt repository.
> > >
> > > Regards
> > >
> > > Owen
> > >
> >
> > --
> > ========================================================================
> > Dr Greig A Cowan http://www.ph.ed.ac.uk/~gcowan1
> > School of Physics, University of Edinburgh, James Clerk Maxwell Building
> >
> > TIER-2 STORAGE SUPPORT PAGES: http://wiki.gridpp.ac.uk/wiki/Grid_Storage
> > ========================================================================
>
--
=======================================================================
Dr Greig A Cowan http://www.ph.ed.ac.uk/~gcowan1
School of Physics, University of Edinburgh, James Clerk Maxwell Building
TIER-2 STORAGE SUPPORT PAGES: http://wiki.gridpp.ac.uk/wiki/Grid_Storage
=======================================================================
|