r27 - 02 Mar 2010 - 12:19:46 - AlvaroFernandezYou are here: TWiki > 
Atlas Web > GridComputing > EGEEHome > LustreStoRM
Atlas Web > GridComputing > EGEEHome > LustreStoRMLustre and StoRM?
Lustre Configuration and Setup
Software Installation.
O.S: Headnode and OSS with Scientific Linux IFIC 4.4 (x86_64)Client with Scientific Linux IFIC 3.0.6 (i386) Updated software in sli44 machines:
acl-2.2.39-1.1.x86_64.rpm libacl-2.2.39-1.1.x86_64.rpm attr-2.4.28-1.2.x86_64.rpm libattr-2.4.28-1.2.x86_64.rpm dump-0.4b41-1.x86_64.rpm rmt-0.4b41-1.x86_64.rpm e2fsprogs-1.39.cfs2-0.x86_64.rpm (the e2fsprogs version must be equal or higher than 1.38)Updated/Installed/Removed software in sli306 machine:
rpm -Uhv libgcc-3.4.6-3.i386.rpm rpm -ihv libstdc++-3.4.6-3.i386.rpm rpm -ihv db4-4.2.52-7.1.i386.rpm rpm -Uhv e2fsprogs-1.39.cfs2-0.i386.rpm rpm -e modutils rpm -ihv module-init-tools-3.1-0.pre5.3.2.i386.rpmLustre: kernel, modules and lustre utilities.
kernel-lustre-smp-2.6.9-42.0.10.EL_lustre_1.6.0.1.x86_64.rpm lustre-modules-1.6.0.1-2.6.9_42.0.10.EL_lustre_1.6.0.1smp.x86_64.rpm lustre-1.6.0.1-2.6.9_42.0.10.EL_lustre_1.6.0.1smp.x86_64.rpm
Setup
Sincronize nodes with ntp. Disable selinux until create a proper profileNodes setup
Headnode: will host MGS + MDT
MGS
mkfs.lustre --mgs /dev/sda5 mkdir -p /mnt/lustre/mgs mount -t lustre /dev/sda5 /mnt/lustre/mgs
MDT
mkfs.lustre --fsname=ificfs --mdt --mgsnode=wn181@tcp0 --mountfsoptions=acl /dev/sda6 mkdir -p /mnt/lustre/mdt-ificfs mount /dev/sda6 /mnt/lustre/mdt-ificfsLook for the partition labels with
e2label and add the lines to fstab for next mounts.
LABEL=MGS /mnt/lustre/mgs lustre defaults,_netdev 0 0 LABEL=ificfs-MDT0000 /mnt/lustre/mdt-ificfs lustre defaults,_netdev 0 0
OSS:
mkfs.lustre --fsname=ificfs --ost --mgsnode=wn181@tcp0 /dev/sda5 mkdir -p /mnt/lustre/ost0-ificfs mount -t lustre /dev/sda5 /mnt/lustre/ost0-ificfs/fstab entry:
LABEL=ificfs-OST0000 /mnt/lustre/ost0-ificfs lustre defaults,_netdev 0 0
Client:
mkdir -p /lustre/ific.uv.es/ mount -t lustre wn181.ific.uv.es:/ificfs /lustre/ific.uv.es/fstab entry:
wn181@tcp:/ificfs /lustre/ific.uv.es lustre defaults,_netdev 0 0
StoRM Configuration and Setup
Current services
The current installation in production is based on Storm release 1.4.0.We are currently evaluating version 1.5 to be installed first in the pre-production services: http://igrelease.forge.cnaf.infn.it/doku.php?id=doc:updates:ig3_1_32:infngrid-update31-61
OS, Required packages & Storm Installation
O.S: Storm is currently released for Scientific Linux 4, and whe are using SLI 4.7, which is the Cern Version with some IFIC packages and configurations for our institute.Relases for SL5 is promised by the developers, and although there are some users that have installed it over a SL5 there are a high dependency to legacy packages that makes it nonsense for us to install it in this moment.
Repositories: We are using yum for the os, base and middleware repositories.
-rw-r--r-- 1 root root 256 Nov 13 12:03 ific-extra.repo -rw-r--r-- 1 root root 585 Nov 13 12:03 glite-generic.repo -rw-r--r-- 1 root root 362 Nov 13 12:03 dag.repo -rw-r--r-- 1 root root 279 Nov 13 12:03 lcg-CA.repo -rw-r--r-- 1 root root 630 Nov 13 12:03 jpackage.repo -rw-r--r-- 1 root root 330 Nov 13 12:03 ig.repo -rw-r--r-- 1 root root 350 Nov 13 12:03 ific-update-srpms.repo -rw-r--r-- 1 root root 328 Nov 13 12:03 ific-update.repo -rw-r--r-- 1 root root 475 Nov 13 12:03 ific-test-srpms.repo.rpmorig -rw-r--r-- 1 root root 244 Nov 13 12:03 ific-srpms.repo -rw-r--r-- 1 root root 222 Nov 13 12:03 ific.repo -rw-r--r-- 1 root root 279 Nov 13 12:03 ific-extra-srpms.repo -rw-r--r-- 1 root root 225 Nov 13 12:38 ific4x.repo -rw-r--r-- 1 root root 331 Nov 13 12:39 ific-update-4x.repoNTP: Install and configure ntp for node date syncronization Java: Installed Java 1_5_0_15 and exclude jre from updates (/etc/yum.conf).
java-1.5.0-sun-devel-1.5.0.15-1jpp java-1.5.0-sun-1.5.0.15-1jppIf there are dependency problems in the installatin you can remove the packages that require jre, in our SLI installation:
mozilla-plugin, ificdummy openoffice* jere susefax susefax-coverInstall CAs and Storm packages:
yum install lcg-Ca yum install -y ig_SE_storm_backend ig_SE_storm_frontendwe are using another node/s for GridFTP? so we don't install GridFtp? here although this is a possible scenario for smaller sites. Depending on acess you can also install frontend and backend in different nodes, or even having several frontends for larger sites. Other: copy grid certificates /etc/grid-security In production we create the user "storm" (with id 666:666 ) prior to yaim configuration to maintain the permissions to access the data in /Lustre.
Configuration
We are using the ig-yaim release provided by INFN wich builds in top of generic gLite releases. Following instructions as indicated in http://igrelease.forge.cnaf.infn.it/doku.php?id=doc:guides:install-3_1 Yaim packages:Ig-yaim-storm-4.0.8-1 glite-yaim-core-4.0.8-7 ig-yaim-4.0.8-3_1Prepare the configuration files: Yaim is able to maintain a site config file, and a per-service file to maintain the variables belonging to this service. We include here the specific Storm variables in the file services/ig-se_storm_backend (*check below some notes about not included parameters)
######################################### # StoRM backend configuration variables # ######################################### #ig : necesario para igyaim-config NTP_HOSTS_IP=147.156.1.1 STORM_HOST=$SE_SRM_HOST STORM_PORT=8443 STORM_USER=storm STORM_GRIDFTP_POOL_LIST=(tux4u01.ific.uv.es) # Database settings. # Host for database connection. COMPULSORY STORM_DB_HOST=localhost # User for database connection. COMPULSORY STORM_DB_USER=storm # Protocol support. # If set to 'FALSE', the following variables prevent the corresponding protocol # to be published by the StoRM gip. OPTIONAL - Available values: [true|false] - Default value: true #STORM_INFO_FILE_SUPPORT=true #STORM_INFO_GRIDFTP_SUPPORT=true #STORM_INFO_RFIO_SUPPORT=true #STORM_INFO_ROOT_SUPPORT=true STORM_INFO_RFIO_SUPPORT=false STORM_INFO_ROOT_SUPPORT=false STORM_DEFAULT_ROOT=/lustre/ific.uv.es/grid ATLAS_STORAGE_AREAS="atlas atlasdatadisk atlasmcdisk atlasenduser atlasgroupdisk atlashotdisk atlaslocalgroupdisk atlasproddisk atlasscratchdisk atlasuserdisk" STORM_STORAGEAREA_LIST=" $ATLAS_STORAGE_AREAS dteam ops swetest ific opscsic gencsic ngies opsngi genngi formngi grid4build nanodev mosfet archist blast slgrid photonics qcomp frodock odthpiv filogen timones gphase turbulencia meteo" STORM_ATLASDATADISK_VONAME="atlas" STORM_ATLASDATADISK_ROOT=$STORM_DEFAULT_ROOT/atlas/atlasdatadisk STORM_ATLASDATADISK_TOKEN="ATLASDATADISK" STORM_ATLASDATADISK_ACCESSPOINT=$STORM_ATLASDATADISK_ROOT STORM_ATLASMCDISK_VONAME="atlas" STORM_ATLASMCDISK_ROOT=$STORM_DEFAULT_ROOT/atlas/atlasmcdisk STORM_ATLASMCDISK_TOKEN="ATLASMCDISK" STORM_ATLASMCDISK_ACCESSPOINT=$STORM_ATLASMCDISK_ROOT STORM_ATLASENDUSER_VONAME="atlas" STORM_ATLASENDUSER_ROOT=$STORM_DEFAULT_ROOT/atlas/atlasenduser STORM_ATLASENDUSER_TOKEN="ATLASENDUSER" STORM_ATLASENDUSER_ACCESSPOINT=$STORM_ATLASENDUSER_ROOT STORM_ATLASGROUPDISK_VONAME="atlas" STORM_ATLASGROUPDISK_ROOT=$STORM_DEFAULT_ROOT/atlas/atlasgroupdisk STORM_ATLASGROUPDISK_TOKEN="ATLASGROUPDISK" STORM_ATLASGROUPDISK_ACCESSPOINT=$STORM_ATLASGROUPDISK_ROOT STORM_ATLASHOTDISK_VONAME="atlas" STORM_ATLASHOTDISK_ROOT=$STORM_DEFAULT_ROOT/atlas/atlashotdisk STORM_ATLASHOTDISK_TOKEN="ATLASHOTDISK" STORM_ATLASHOTDISK_ACCESSPOINT=$STORM_ATLASHOTDISK_ROOT STORM_ATLASLOCALGROUPDISK_VONAME="atlas" STORM_ATLASLOCALGROUPDISK_ROOT=$STORM_DEFAULT_ROOT/atlas/atlaslocalgroupdisk STORM_ATLASLOCALGROUPDISK_TOKEN="ATLASLOCALGROUPDISK" STORM_ATLASLOCALGROUPDISK_ACCESSPOINT=$STORM_ATLASLOCALGROUPDISK_ROOT STORM_ATLASPRODDISK_VONAME="atlas" STORM_ATLASPRODDISK_ROOT=$STORM_DEFAULT_ROOT/atlas/atlasproddisk STORM_ATLASPRODDISK_TOKEN="ATLASPRODDISK" STORM_ATLASPRODDISK_ACCESSPOINT=$STORM_ATLASPRODDISK_ROOT STORM_ATLASSCRATCHDISK_VONAME="atlas" STORM_ATLASSCRATCHDISK_ROOT=$STORM_DEFAULT_ROOT/atlas/atlasscratchdisk STORM_ATLASSCRATCHDISK_TOKEN="ATLASSCRATCHDISK" STORM_ATLASSCRATCHDISK_ACCESSPOINT=$STORM_ATLASSCRATCHDISK_ROOT STORM_ATLASUSERDISK_VONAME="atlas" STORM_ATLASUSERDISK_ROOT=$STORM_DEFAULT_ROOT/atlas/atlasuserdisk STORM_ATLASUSERDISK_TOKEN="ATLASUSERDISK" STORM_ATLASUSERDISK_ACCESSPOINT=$STORM_ATLASUSERDISK_ROOT STORM_ATLAS_VONAME="atlas" STORM_ATLAS_ROOT=$STORM_ATLASSCRATCHDISK_ROOT STORM_ATLAS_TOKEN="ATLAS" STORM_ATLAS_ACCESSPOINT=$STORM_ATLAS_ROOTonly non-default paramenters have been included here, the rest are #commented and not included here for the sake of simplicity. About the defined Storage areas, are defined in the STORM_STORAGEAREA_LIST variable and configured in the STORM_<VO_NAME>_* related variables. But only the ones refering to ATLAS tokens are included here since the others are similar are ATLAS are in the scope of this wiki. Configuration files issue: the infn release installs by default several services that we don't need. We remove the configuration for GRIDEYE and NFS ( commenting/removing corresponding entry in yaim file: /opt/glite/yaim/node-info.d/ig-se_storm_backend):
/opt/glite/yaim/node-info.d/ig-se_storm_backend: #config_fmon_client #config_nfs_sw_dir_serverConfigure the services:
/opt/glite/yaim/bin/ig_yaim -c -s site-info-pps-ific-SL4.def -n ig_SE_storm_backend -n ig_SE_storm_frontend
ATLAS authorization configuration
To allow a proper authorization configuration an being able access to data, we are using a special authorization plugin developed by us, and that now it is included in the main storm sources. This means you can use it, but it is not very well documented.
Modify the plugin to be used, and related variables in the storm.properties file:
/opt/storm/backend/etc/storm.properties:
authorization.sources = LocalAuthorizationSource
# =======================
# nuevas directrices
# =======================
# poner el permiso de escritura en directorios nuevos
# lo necesita LocalAuthorizationSource
directory.writeperm = true
In our srm and gridftp configuration, all atlas users all mapped to a single user. Check the mapping files:
/etc/grid-security/voms-grid-mapfile:
"/atlas/Role=lcgadmin" atls000
"/atlas/Role=lcgadmin/Capability=NULL" atls000
"/atlas/Role=production" atlp000
"/atlas/Role=production/Capability=NULL" atlp000
"/atlas/Role=software" atls000
"/atlas/Role=software/Capability=NULL" atls000
"/atlas/Role=pilot" atlu000
"/atlas/Role=pilot/Capability=NULL" atlu000
"/atlas" atlu000
"/atlas/Role=NULL" atlu000
"/atlas/Role=NULL/Capability=NULL" atlu000
file /opt/edg/etc/edg-mkgridmap.conf:
# ATLAS
# Map VO members (prd)
group vomss://voms.cern.ch:8443/voms/atlas?/atlas/Role=production atlp000
# Map VO members (sgm)
group vomss://voms.cern.ch:8443/voms/atlas?/atlas/Role=lcgadmin atls000
# Map VO members (root Group)
group vomss://voms.cern.ch:8443/voms/atlas?/atlas atlu000
The idea is that with this configuration the plugin respects the ACLs on disk to authorize or not permission to access the files, according to ATLAS policies. You can check the ACLs that we have defined in the following spaces of the atlas pool in /lustre/ific.uv.es/grid/atlas:
# file: .. # owner: root # group: root user::rwx group::r-x other::r-x # file: . # owner: atlu000 # group: atlas user::rwx user:101:rwx group::rwx mask::rwx other::r-x # file: atlasdatadisk # owner: storm # group: storm user::rwx group::r-x group:atlas:r-x group:atlp:rwx mask::rwx other::--- default:user::rwx default:group::r-x default:group:atlas:r-x default:mask::rwx default:other::--- # file: atlasenduser # owner: storm # group: storm user::rwx group::rwx group:atlas:rwx mask::rwx other::--- # file: atlasgroupdisk # owner: storm # group: storm user::rwx group::r-x group:atlas:r-x group:atlp:rwx mask::rwx other::--- default:user::rwx default:group::r-x default:group:atlas:r-x default:mask::rwx default:other::--- # file: atlashotdisk # owner: storm # group: storm user::rwx group::r-x group:atlas:r-x group:atlp:rwx group:atls:--x mask::rwx other::--- default:user::rwx default:group::r-x default:group:atlas:r-x default:mask::rwx default:other::r-x # file: atlaslocalgroupdisk # owner: storm # group: storm user::rwx group::rwx group:atlas:rwx group:atlp:--x mask::rwx other::--- # file: atlasmcdisk # owner: storm # group: storm user::rwx group::r-x group:atlas:r-x group:atlp:rwx mask::rwx other::--- default:user::rwx default:group::r-x default:group:atlas:r-x default:mask::rwx default:other::--- # file: atlasproddisk # owner: storm # group: storm user::rwx group::r-x group:atlas:r-x group:atlp:rwx mask::rwx other::--- default:user::rwx default:group::r-x default:group:atlas:r-x default:mask::rwx default:other::--- # file: atlasscratchdisk # owner: storm # group: storm user::rwx group::rwx group:atlas:rwx group:atlp:--x mask::rwx other::--- # file: atlasuserdisk # owner: storm # group: storm user::rwx group::rwx group:atlas:rwx group:atlp:--x group:atls:--x mask::rwx other::--- # file: generated # owner: atlu000 # group: atlas user::rwx group::rwx other::r-xSee in the previous box for example the atlasproddisk space, where common atlas users can access and read the files, but only production users (atlp mapped GID) can write. The access ACLs are created for this policy, and the default acls and mask to respect it in the created files and subdirectories.
Dinamic Info Provider
The initial default configuration for the available and free space in disk checks the disk, but it does not take into account our pool configuration for the different spaces, specially for ATLAS vo. It does not update correctly the information when it is runnint, in the Information system and in the internal Storm database. To be able to publish information correctly we developed a script to update this data frequently, that you can find here: space-tokens-dynamicinfo_v2.tgzConfiguration issues
- If you are not using the default port, check the configuration file since yaim is not changing it properly
/opt/storm/backend/etc/storm.properties:
storm.service.port=8444
fe.port = 8444
- On previous version configuration was not understanding correctly FQAN at voms proxies, like this:
/opt/srmv2storm/var/log/srmv2storm.log: 06/14 18:15:46 26838,0 Rm: UserDN=/C=ES/O=DATAGRID-ES/O=IFIC/CN=Alvaro Fernandez Casani 06/14 18:15:46 26838,0 Rm: Number of FQANs: 0
- in this case, check that the yaim configuration creates the correct directories containing the voms servers certificates ( or for newer versions the DN of the voms server ):
[root@srmv2 storm]# ll /etc/grid-security/vomsdir/ total 432 drwxr-xr-x 2 root root 4096 Nov 26 15:12 atlas -rw-r--r-- 1 root root 4620 Jan 19 15:41 cclcgvomsli01.in2p3.fr.1413.pem -rw-r--r-- 1 root root 1440 Dec 2 16:09 cert-voms-01.cnaf.infn.it.pem -rw-r--r-- 1 root root 1440 Dec 2 16:09 cert-voms-01.cnaf.infn.it.pem.1 -rw-r--r-- 1 root root 1424 Dec 2 16:09 cert-voms-01.cnaf.infn.it.pem.2 -rw-r--r-- 1 root root 5857 Mar 17 2009 dgrid-voms.fzk.de.4515.pem drwxr-xr-x 2 root root 4096 Nov 13 13:15 dteam -rw-r--r-- 1 root root 5182 Mar 17 2009 glite-io.scai.fraunhofer.de.4499.pem -rw-r--r-- 1 root root 1436 Dec 2 16:09 grid12.lal.in2p3.fr.pem -rw-r--r-- 1 root root 5132 Mar 17 2009 grid-voms.desy.de.5369.pem -rw-r--r-- 1 root root 5132 Mar 17 2009 grid-voms.desy.de.8119.pem -rw-r--r-- 1 root root 1428 Mar 17 2009 grid-voms.esrf.eu.4140.pem drwxr-xr-x 2 root root 4096 Nov 13 13:15 ific -rw-r--r-- 1 root root 5968 Jan 19 15:41 lcg-voms.cern.ch.2009-03-03.pem -rw-r--r-- 1 root root 6782 Jan 19 15:41 lcg-voms.cern.ch.2010-01-18.pem -rw-r--r-- 1 root root 5154 Dec 2 16:09 mu4.matrix.sara.nl.pem drwxr-xr-x 2 root root 4096 Nov 13 13:15 ngi-es drwxr-xr-x 2 root root 4096 Nov 23 22:40 ops drwxr-xr-x 2 root root 4096 Nov 13 13:15 swetest -rw-r--r-- 1 root root 2496 Nov 27 13:57 swevo.ific.uv.es.pem drwxr-xr-x 2 root root 4096 Nov 13 13:15 vo.archist.es-ngi.eu drwxr-xr-x 2 root root 4096 Nov 13 13:15 vo.blast.es-ngi.eu drwxr-xr-x 2 root root 4096 Nov 13 13:15 vo.filogen.es-ngi.eu drwxr-xr-x 2 root root 4096 Nov 13 13:15 vo.formacion.es-ngi.eu drwxr-xr-x 2 root root 4096 Nov 13 13:15 vo.frodock.es-ngi.eu drwxr-xr-x 2 root root 4096 Nov 13 13:15 vo.general.csic.es drwxr-xr-x 2 root root 4096 Nov 13 13:15 vo.general.es-ngi.eu drwxr-xr-x 2 root root 4096 Nov 13 13:15 vo.gphase.es-ngi.eu drwxr-xr-x 2 root root 4096 Nov 13 13:15 vo.grid4build.es-ngi.eu drwxr-xr-x 2 root root 4096 Nov 13 13:15 vo.meteo.es-ngi.eu drwxr-xr-x 2 root root 4096 Nov 13 13:15 vo.mosfet.es-ngi.eu -rw-r--r-- 1 root root 2502 Jul 7 2009 voms01.ifca.es.pem -rw-r--r-- 1 root root 1419 Dec 2 16:09 voms-01.pd.infn.it.pem -rw-r--r-- 1 root root 1420 Dec 2 16:09 voms-01.pd.infn.it.pem.1 -rw-r--r-- 1 root root 1419 Dec 2 16:09 voms-02.pd.infn.it.pem -rw-r--r-- 1 root root 1420 Dec 2 16:09 voms-02.pd.infn.it.pem.1 -rw-r--r-- 1 root root 1424 Dec 2 16:09 voms2.cnaf.infn.it.pem -rw-r--r-- 1 root root 1424 Dec 2 16:09 voms2.cnaf.infn.it.pem.1 -rw-r--r-- 1 root root 1404 Dec 2 16:09 voms2.cnaf.infn.it.pem.2 -rw-r--r-- 1 root root 6552 Jan 19 15:41 voms.cern.ch.2009-06-22.pem -rw-r--r-- 1 root root 1419 Dec 2 16:09 voms.cnaf.infn.it.pem -rw-r--r-- 1 root root 1419 Dec 2 16:09 voms.cnaf.infn.it.pem.1 -rw-r--r-- 1 root root 1399 Dec 2 16:09 voms.cnaf.infn.it.pem.2 -rw-r--r-- 1 root root 4865 Jan 19 15:41 voms.fnal.gov.35501.pem -rw-r--r-- 1 root root 1484 Dec 2 16:09 voms.fnal.gov.pem -rw-r--r-- 1 root root 1298 Dec 2 16:09 voms.fnal.gov.pem.1 -rw-r--r-- 1 root root 1843 Dec 2 16:09 voms.gridpp.ac.uk.pem -rw-r--r-- 1 root root 2138 Dec 2 16:09 voms.gridpp.ac.uk.pem.1 -rw-r--r-- 1 root root 1651 Dec 2 16:09 voms.grid.sara.nl.pem -rw-r--r-- 1 root root 5152 Dec 2 16:09 voms.grid.sara.nl.pem.1 -rw-r--r-- 1 root root 2483 Nov 27 14:05 voms.ific.uv.es.pem -rw-r--r-- 1 root root 6566 Jan 19 15:41 voms-pilot.cern.ch.2009-06-30.pem -rw-r--r-- 1 root root 1472 Dec 2 16:09 voms.research-infrastructures.eu.pem -rw-r--r-- 1 root root 1472 Dec 2 16:09 voms.research-infrastructures.eu.pem.1 drwxr-xr-x 2 root root 4096 Nov 13 13:15 vo.nanodev.es-ngi.eu drwxr-xr-x 2 root root 4096 Nov 13 13:15 vo.odthpiv.es-ngi.eu drwxr-xr-x 2 root root 4096 Nov 13 13:15 vo.operaciones.es-ngi.eu drwxr-xr-x 2 root root 4096 Nov 26 15:12 vo.ops.csic.es drwxr-xr-x 2 root root 4096 Nov 13 13:15 vo.photonics.es-ngi.eu drwxr-xr-x 2 root root 4096 Nov 13 13:15 vo.qcomp.es-ngi.eu -rw-r--r-- 1 root root 4697 Jan 19 15:41 vo.racf.bnl.gov.40260.pem drwxr-xr-x 2 root root 4096 Nov 13 13:15 vo.slgrid.es-ngi.eu drwxr-xr-x 2 root root 4096 Nov 13 13:15 vo.timones.es-ngi.eu drwxr-xr-x 2 root root 4096 Nov 13 13:15 vo.turbulencia.es-ngi.eu
- Then correct interpretation of FQANs is acomplished:
02/18 13:13:08 495,0 Mkdir: UserDN=/DC=es/DC=irisgrid/O=ific/CN=Alvaro-Fernandez 02/18 13:13:08 495,0 Mkdir: Number of FQANs: 2 02/18 13:13:08 495,0 Mkdir: FQAN[0]: /atlas/Role=NULL/Capability=NULL 02/18 13:13:08 495,0 Mkdir: FQAN[1]: /atlas/lcg1/Role=NULL/Capability=NULL
- If there are problems with java in reconfiguration, then it gives problem locating java and starting storm-backend server. Solution:
[root@ccc01 yaim]# alternatives --install /usr/bin/java java /usr/java/jdk1.5.0_07/bin/java 1 [root@ccc01 yaim]# alternatives --config java There are 2 programs which provide 'java'. Selection Command ----------------------------------------------- *+ 1 /usr/share/java/libgcj-java-placeholder.sh 2 /usr/java/jdk1.5.0_07/bin/java Enter to keep the current selection[+], or type selection number: 2
Srm command testing
We developed a simple battery of test to check the srm functionality of the installed storm server. You can find the results below and the test script here: test-srm.sh You will need the srm client that is distributed in the Storm webpage: http://storm.forge.cnaf.infn.it/downloads/clientsrmPing
Testing Ping command: ./clientSRM Ping -e httpg://srmv2.ific.uv.es:8443
============================================================
Sending Ping request to: httpg://srmv2.ific.uv.es:8443
============================================================
Request status:
statusCode="SRM_SUCCESS"(0)
explanation="SRM server successfully contacted"
============================================================
SRM Response:
versionInfo="v2.2"
otherInfo (size=2)
[0] key="backend_type"
[0] value="StoRM"
[1] key="backend_version"
[1] value="<FE:1.4.0-01.sl4><BE:1.4.0-00>"
============================================================
Ping Command:
[ OK ]
srmMkdir
Testing Mkdir command: ./clientSRM mkdir -e httpg://srmv2.ific.uv.es:8443 -s srm://srmv2.ific.uv.es:8443//lustre/ific.uv.es/grid/atlas/atlasscratchdisk/test_mkdir_26999
============================================================
Sending Mkdir request to: httpg://srmv2.ific.uv.es:8443
============================================================
Request status:
statusCode="SRM_SUCCESS"(0)
explanation="Directory created with success"
============================================================
SRM Response:
============================================================
Mkdir Command:
[ OK ]
srmRmdir
Testing Rmdir command: ./clientSRM rmdir -e httpg://srmv2.ific.uv.es:8443 -s srm://srmv2.ific.uv.es:8443//lustre/ific.uv.es/grid/atlas/atlasscratchdisk/test_mkdir_26999
============================================================
Sending Rmdir request to: httpg://srmv2.ific.uv.es:8443
============================================================
Request status:
statusCode="SRM_SUCCESS"(0)
explanation="Directory removed with success!"
============================================================
SRM Response:
============================================================
Rmdir Command:
[ OK ]
srmReserveSpace
Testing ReserveSpace command: ./clientSRM ReserveSpace -e httpg://srmv2.ific.uv.es:8443 -a 10000000 -b 5000000 -r 0,0
============================================================
Sending ReserveSpace request to: httpg://srmv2.ific.uv.es:8443
============================================================
Request status:
statusCode="SRM_SUCCESS"(0)
explanation="Space Reservation done"
============================================================
SRM Response:
sizeOfTotalReservedSpace=10000000
sizeOfGuaranteedReservedSpace=10000000
lifetimeOfReservedSpace=2147483647
spaceToken="D2EF3295-2B01-1F6D-939C-74E8AF1C71F6"
============================================================
ReserveSpace Command:
[ OK ]
where parameters are the following: - -a <desiredSizeOfTotalSpace> in millions of bytes (i.e. 10Mb )
- -b <desiredSizeOfGuaranteedSpace> (i.e. 5Mb)
- -r <retentionPolicy,accessLatency> Retention policy (0=replica, 1=output, 2=custodial),(0=online, 1=nearline)
Currently StoRM? supports:
* Retention Policy = { REPLICA }
* Access Latency = { ONLINE }
* ExpirationMode? = { releaseWhenExpire, neverExpire } (Note: neverExpire will be the default)
* Access Protocol = { gsiftp, file, rfio }
*Access Pattern = { WAN, LAN } (Note: It depends on the access protocol)
srmCopy
*note*:srmCopy is working only with SURL and not with local file. This is not reflected in the documentation
[alferca@cg02 srmtest]$ ./clientSRM copy -e httpg://ccc01.ific.uv.es:8444/ -s file:///tmp/kk.log srm://ccc01.ific.uv.es:8444/atlas/filekk
============================================================
Sending Copy request to: httpg://ccc01.ific.uv.es:8444/
============================================================
Request status:
statusCode="SRM_REQUEST_QUEUED"(17)
explanation=""
============================================================
SRM Response:
requestToken="59cbef94-a552-4886-96af-e55dae60c144"
arrayOfFileStatuses (size=1)
[0] sourceSURL="file:///tmp/kk.log"
[0] targetSURL="srm://ccc01.ific.uv.es:8444/atlas/filekk"
[0] status: statusCode="SRM_REQUEST_QUEUED"(17)
explanation=""
============================================================
[alferca@cg02 srmtest]$ ./clientSRM statuscopy -e httpg://ccc01.ific.uv.es:8444/ -t 59cbef94-a552-4886-96af-e55dae60c144
============================================================
Sending StatusCopy request to: httpg://ccc01.ific.uv.es:8444/
============================================================
Request status:
statusCode="SRM_FAILURE"(1)
explanation="This SRM Copy request contained nothing to process!"
============================================================
SRM Response:
remainingTotalRequestTime=0
arrayOfFileStatuses (size=1)
[0] sourceSURL="file:///tmp/kk.log"
[0] targetSURL="srm://ccc01.ific.uv.es:8444/atlas/filekk"
[0] estimatedWaitTime=-1
[0] status: statusCode="SRM_FAILURE"(1)
explanation="This SRM Copy request contained nothing to process!"
============================================================
srmCopy in pull mode is not implemente, and this tests currently *FAILS*. Actually if you want add a local file to a storage through an SRM we suggest to use the sequence
of operation:
- prepareToPut with the target SURL
- globus-url-copy with the TURL returned by the previous srmPrepareToGet
- srmPutDone to notify to the SRM the end of transfer.
SrmPtP?
Testing Ptp command: ./clientSRM Ptp -e httpg://srmv2.ific.uv.es:8443 -s file:////tmp/testfile.26999 srm://srmv2.ific.uv.es:8443//lustre/ific.uv.es/grid/atlas/atlasscratchdisk/test_file_26999
============================================================
Sending PtP request to: httpg://srmv2.ific.uv.es:8443
============================================================
Request status:
statusCode="SRM_REQUEST_QUEUED"(17)
explanation=""
============================================================
SRM Response:
requestToken="ff4a0ea8-07d4-498b-a37b-847a7f530f3a"
arrayOfFileStatuses (size=2)
[0] SURL="file:////tmp/testfile.26999"
[0] status: statusCode="SRM_REQUEST_QUEUED"(17)
explanation=""
[1] SURL="srm://srmv2.ific.uv.es:8443//lustre/ific.uv.es/grid/atlas/atlasscratchdisk/test_file_26999"
[1] status: statusCode="SRM_REQUEST_QUEUED"(17)
explanation=""
============================================================
Ptp Command:
[ OK ]
SrmStatusPtP?
Testing StatusPtp command: ./clientSRM StatusPtp -e httpg://srmv2.ific.uv.es:8443 -t ff4a0ea8-07d4-498b-a37b-847a7f530f3a
============================================================
Sending StatusPtP request to: httpg://srmv2.ific.uv.es:8443
============================================================
Request status:
statusCode="SRM_SUCCESS"(0)
explanation="All chunks successfully handled!"
============================================================
SRM Response:
remainingTotalRequestTime=0
arrayOfFileStatuses (size=2)
[0] SURL="file:////tmp/testfile.26999"
[0] status: statusCode="SRM_FAILURE"(1)
explanation="This chunk of the request is malformed!"
[0] estimatedWaitTime=-1
[1] SURL="srm://srmv2.ific.uv.es:8443//lustre/ific.uv.es/grid/atlas/atlasscratchdisk/test_file_26999"
[1] status: statusCode="SRM_SPACE_AVAILABLE"(24)
explanation="srmPrepareToPut successfully handled!"
[1] estimatedWaitTime=-1
[1] TURL="gsiftp://tux4u04.ific.uv.es:2811//lustre/ific.uv.es/grid/atlas/atlasscratchdisk/test_file_26999"
============================================================
StatusPtp Command:
[ OK ]
and when StatusPtP? is OK, then the data can be send with gridftp:
globus-url-copy file:///tmp/testfile.26999 gsiftp://tux4u04.ific.uv.es:2811//lustre/ific.uv.es/grid/atlas/atlasscratchdisk/test_file_26999 Transfered file with globus-url-copy 1024+0 records in [ OK ] 1024+0 records out
SrmPutDone?
Testing PutDone command: ./clientSRM PutDone -e httpg://srmv2.ific.uv.es:8443 -s srm://srmv2.ific.uv.es:8443//lustre/ific.uv.es/grid/atlas/atlasscratchdisk/test_file_26999 -t ff4a0ea8-07d4-498b-a37b-847a7f530f3a
============================================================
Sending PutDone request to: httpg://srmv2.ific.uv.es:8443
============================================================
Request status:
statusCode="SRM_SUCCESS"(0)
explanation="All file requests are successfully completed"
============================================================
SRM Response:
arrayOfFileStatuses (size=1)
[0] SURL="srm://srmv2.ific.uv.es:8443/lustre/ific.uv.es/grid/atlas/atlasscratchdisk/test_file_26999"
[0] status: statusCode="SRM_SUCCESS"(0)
explanation=""
============================================================
PutDone Command:
[ OK ]
SrmPrepareToGet?
Testing PtG command: ./clientSRM PtG -p -e httpg://srmv2.ific.uv.es:8443 -s srm://srmv2.ific.uv.es:8443//lustre/ific.uv.es/grid/atlas/atlasscratchdisk/test_file_26999 -T -P file
============================================================
Sending PtG request to: httpg://srmv2.ific.uv.es:8443
============================================================
Polling request status:
Current status: SRM_REQUEST_QUEUED (Ctrl+c to stop polling).
============================================================
Request status:
statusCode="SRM_SUCCESS"(0)
explanation="All chunks successfully handled!"
============================================================
SRM Response:
requestToken="e11bd7aa-5136-4d7a-b740-be2dd6a0c3b6"
remainingTotalRequestTime=0
arrayOfFileStatuses (size=1)
[0] status: statusCode="SRM_FILE_PINNED"(22)
explanation="srmPrepareToGet successfully handled!"
[0] sourceSURL="srm://srmv2.ific.uv.es:8443//lustre/ific.uv.es/grid/atlas/atlasscratchdisk/test_file_26999"
[0] fileSize=1048576
[0] estimatedWaitTime=-1
[0] transferURL="file:///lustre/ific.uv.es/grid/atlas/atlasscratchdisk/test_file_26999"
============================================================
PtG Command:
[ OK ]
SrmGetSpaceTokens?
Testing GetSpaceTokens command: ./clientSRM GetSpaceTokens -v -e httpg://srmv2.ific.uv.es:8443 -d ATLASMCDISK
============================================================
Sending GetSpaceTokens request to: httpg://srmv2.ific.uv.es:8443
============================================================
Request status:
statusCode="SRM_SUCCESS"(0)
explanation=""
============================================================
SRM Response:
arrayOfSpaceTokens(size=1)
[0] "1405D063-2901-073C-939C-74E8BBCFCD45"
============================================================
GetSpaceTokens Command:
[ OK ]
SrmGetSpaceMetadata?
Testing GetSpaceMetadata command: ./clientSRM GetSpaceMetadata -v -e httpg://srmv2.ific.uv.es:8443 -s
1405D063-2901-073C-939C-74E8BBCFCD45
============================================================
Sending GetSpaceMetaData request to: httpg://srmv2.ific.uv.es:8443
============================================================
Request status:
statusCode="SRM_SUCCESS"(0)
explanation=""
============================================================
SRM Response:
arrayOfSpaceDetails (size=1)
[0] spaceToken="1405D063-2901-073C-939C-74E8BBCFCD45"
[0] status: statusCode="SRM_SUCCESS"(0)
explanation="Valid space token"
[0] owner="/DC=it/DC=infngrid/OU=Services/CN=storm"
[0] totalSize=154568936964096
[0] guaranteedSize=154568936964096
[0] unusedSize=60829260051832
[0] lifetimeAssigned=0
[0] lifetimeLeft=0
============================================================
GetSpaceMetadata Command:
1024+0 records in [ OK ]
1024+0 records out
Testing complete cicle to put File
Testing Ptp with space token command: ./clientSRM Ptp -e httpg://srmv2.ific.uv.es:8443 -p -t
1405D063-2901-073C-939C-74E8BBCFCD45 -s file:////tmp/testfile2.26999 srm://srmv2.ific.uv.es:8443//lustre/ific.uv.es/grid/atlas/atlasscratchdisk/test_file2_26999
============================================================
Sending PtP request to: httpg://srmv2.ific.uv.es:8443
============================================================
Polling request status:
Current status: SRM_REQUEST_QUEUED (Ctrl+c to stop polling).
============================================================
Request status:
statusCode="SRM_SUCCESS"(0)
explanation="All chunks successfully handled!"
============================================================
SRM Response:
requestToken="86311065-0014-4e12-9b7f-dc4cdc0850ef"
remainingTotalRequestTime=0
arrayOfFileStatuses (size=2)
[0] SURL="file:////tmp/testfile2.26999"
[0] status: statusCode="SRM_FAILURE"(1)
explanation="This chunk of the request is malformed!"
[0] estimatedWaitTime=-1
[1] SURL="srm://srmv2.ific.uv.es:8443//lustre/ific.uv.es/grid/atlas/atlasscratchdisk/test_file2_26999"
[1] status: statusCode="SRM_SPACE_AVAILABLE"(24)
explanation="srmPrepareToPut successfully handled!"
[1] estimatedWaitTime=-1
[1] TURL="gsiftp://tux4u01.ific.uv.es:2811//lustre/ific.uv.es/grid/atlas/atlasscratchdisk/test_file2_26999"
============================================================
Ptp with space token Command:
[ OK ]
Testing StatusPtp command: ./clientSRM StatusPtp -e httpg://srmv2.ific.uv.es:8443 -t 86311065-0014-4e12-9b7f-dc4cdc0850ef
============================================================
Sending StatusPtP request to: httpg://srmv2.ific.uv.es:8443
============================================================
Request status:
statusCode="SRM_SUCCESS"(0)
explanation="All chunks successfully handled!"
============================================================
SRM Response:
remainingTotalRequestTime=0
arrayOfFileStatuses (size=2)
[0] SURL="file:////tmp/testfile2.26999"
[0] status: statusCode="SRM_FAILURE"(1)
explanation="This chunk of the request is malformed!"
[0] estimatedWaitTime=-1
[1] SURL="srm://srmv2.ific.uv.es:8443//lustre/ific.uv.es/grid/atlas/atlasscratchdisk/test_file2_26999"
[1] status: statusCode="SRM_SPACE_AVAILABLE"(24)
explanation="srmPrepareToPut successfully handled!"
[1] estimatedWaitTime=-1
[1] TURL="gsiftp://tux4u01.ific.uv.es:2811//lustre/ific.uv.es/grid/atlas/atlasscratchdisk/test_file2_26999"
============================================================
StatusPtp Command:
globus-url-copy file:///tmp/testfile2.26999 gsiftp://tux4u01.ific.uv.es:2811//lustre/ific.uv.es/grid/atlas/atlasscratchdisk/test_file2_26999
Transfered file with globus-url-copy:
[ OK ]
Testing PutDone command: ./clientSRM PutDone -e httpg://srmv2.ific.uv.es:8443 -s srm://srmv2.ific.uv.es:8443//lustre/ific.uv.es/grid/atlas/atlasscratchdisk/test_file2_26999 -t 86311065-0014-4e12-9b7f-dc4cdc0850ef
============================================================
Sending PutDone request to: httpg://srmv2.ific.uv.es:8443
============================================================
Request status:
statusCode="SRM_SUCCESS"(0)
explanation="All file requests are successfully completed"
============================================================
SRM Response:
arrayOfFileStatuses (size=1)
[0] SURL="srm://srmv2.ific.uv.es:8443/lustre/ific.uv.es/grid/atlas/atlasscratchdisk/test_file2_26999"
[0] status: statusCode="SRM_SUCCESS"(0)
explanation=""
============================================================
PutDone Command:
1024+0 records in [ OK ]
1024+0 records out
GFAL testing
- The current installation of GFAL and lcg_util tools >= lcg_util-1.5.2 STARTS supporting srmv2.2 protocol.
lcg-cr : copy a file on the remote SE and store it on the catalog
[alferca@cg02 jdl]$ lcg-cr -v --vo atlas -d ccc01.ific.uv.es -l lfn:/grid/atlas/ific.uv.es/test.$$ file:///tmp/kk.log
Using grid catalog type: lfc
Using grid catalog : lfc02.pic.es
Using LFN : /grid/atlas/ific.uv.es/test.12208
Using SURL : srm://ccc01.ific.uv.es/atlas/generated/2007-07-24/file175f15c4-da6a-470c-a8cd-597d2cbd4c88
srm://ccc01.ific.uv.es/atlas/generated/2007-07-24/file175f15c4-da6a-470c-a8cd-597d2cbd4c88: srmPrepareToPut successfully handled!
Source URL: file:///tmp/kk.log
File size: 9
VO name: atlas
Destination specified: ccc01.ific.uv.es
Destination URL for copy: gsiftp://ccc01.ific.uv.es:2811/lustre/ific.uv.es/atlas/generated/2007-07-24/file175f15c4-da6a-470c-a8cd-597d2cbd4c88
# streams: 1
# set timeout to 0 seconds
Alias registered in Catalog: lfn:/grid/atlas/ific.uv.es/test.12208
9 bytes 0.02 KB/sec avg 0.02 KB/sec inst
Transfer took 1060 ms
srm://ccc01.ific.uv.es/atlas/generated/2007-07-24/file175f15c4-da6a-470c-a8cd-597d2cbd4c88:
Destination URL registered in Catalog: srm://ccc01.ific.uv.es/atlas/generated/2007-07-24/file175f15c4-da6a-470c-a8cd-597d2cbd4c88
guid:573a6648-bb4e-4724-ae77-b70141a60946
lcg-cp: copy back a file and store locally
[alferca@cg02 jdl]$ lcg-cp -v --vo atlas lfn:/grid/atlas/ific.uv.es/test.$$ file:///tmp/kk-copied.log
Using grid catalog type: lfc
Using grid catalog : lfc02.pic.es
VO name: atlas
srm://ccc01.ific.uv.es/atlas/generated/2007-07-24/file175f15c4-da6a-470c-a8cd-597d2cbd4c88: srmPrepareToGet successfully handled!
Source URL: lfn:/grid/atlas/ific.uv.es/test.12208
File size: 9
Source URL for copy: gsiftp://ccc01.ific.uv.es:2811/lustre/ific.uv.es/atlas/generated/2007-07-24/file175f15c4-da6a-470c-a8cd-597d2cbd4c88
Destination URL: file:///tmp/kk-copied.log
# streams: 1
# set timeout to 0 (seconds)
0 bytes 0.00 KB/sec avg 0.00 KB/sec inst
Transfer took 1020 ms
Released
lcg-rep: replicate file to a remote storage element
[alferca@cg02 alferca]$ lcg-rep -v --vo atlas -d grid007g.cnaf.infn.it lfn:/grid/atlas/ific.uv.es/test3
Using grid catalog type: lfc
Using grid catalog : lfc02.pic.es
srm://ccc01.ific.uv.es/atlas/generated/2007-07-25/filed5fcf388-ac93-4f2e-a88d-b127eabded18: srmPrepareToGet successfully handled!
Source URL: lfn:/grid/atlas/ific.uv.es/test3
File size: 9
VO name: atlas
Destination specified: grid007g.cnaf.infn.it
Source URL for copy: gsiftp://ccc01.ific.uv.es:2811/lustre/ific.uv.es/atlas/generated/2007-07-25/filed5fcf388-ac93-4f2e-a88d-b127eabded18
Destination URL for copy: gsiftp://grid007g.cnaf.infn.it/flatfiles/SE00/atlas/generated/2007-07-25/filecea8d543-fd6a-4934-be52-0245cd014a7a
# streams: 1
# set timeout to 0
0 bytes 0.00 KB/sec avg 0.00 KB/sec inst
Transfer took 2020 ms
Released
Destination URL registered in LRC: sfn://grid007g.cnaf.infn.it/flatfiles/SE00/atlas/generated/2007-07-25/filecea8d543-fd6a-4934-be52-0245cd014a7a
lcg-lr: list replicas
[alferca@cg02 alferca]$ lcg-lr -v --vo atlas lfn:/grid/atlas/ific.uv.es/test3 sfn://grid007g.cnaf.infn.it/flatfiles/SE00/atlas/generated/2007-07-25/filecea8d543-fd6a-4934-be52-0245cd014a7a srm://ccc01.ific.uv.es/atlas/generated/2007-07-25/filed5fcf388-ac93-4f2e-a88d-b127eabded18
lcg-del: remove replicas of files
[alferca@cg02 alferca]$ lcg-del -v --vo atlas -a lfn:/grid/atlas/ific.uv.es/test3 VO name: atlas Using GUID : 90d0dc51-db18-47c2-9d0c-c6b37e62833e set timeout to 0 seconds sfn://grid007g.cnaf.infn.it/flatfiles/SE00/atlas/generated/2007-07-25/filecea8d543-fd6a-4934-be52-0245cd014a7a is deleted sfn://grid007g.cnaf.infn.it/flatfiles/SE00/atlas/generated/2007-07-25/filecea8d543-fd6a-4934-be52-0245cd014a7a is unregistered srm://ccc01.ific.uv.es/atlas/generated/2007-07-25/filed5fcf388-ac93-4f2e-a88d-b127eabded18: File removed srm://ccc01.ific.uv.es/atlas/generated/2007-07-25/filed5fcf388-ac93-4f2e-a88d-b127eabded18 is deleted srm://ccc01.ific.uv.es/atlas/generated/2007-07-25/filed5fcf388-ac93-4f2e-a88d-b127eabded18 is unregistered
Common Errors
- Authentication errors on MYSQL database, like:
[root@ccc01 current]# /etc/rc.d/init.d/srmv2storm restart srmv2storm already stopped: [FAILED] Starting srmv2storm: bash: /root/.bashrc: Permission denied 07/03 15:16:47 14106 srm_main: verbose level = 4 07/03 15:16:47 14106 srmv2: started 07/03 15:16:47 14106 srmv2: Using 'http://localhost:8080/RPC2' as XMLRPC endpoint 07/03 15:16:47 14106 srmv2: Proxy directory for srmCopy: /opt/storm/var/proxies 07/03 15:16:47 14106 storm_opendb: CONNECT error: Access denied for user 'storm'@'localhost' (using password: YES) 07/03 15:17:47 14106 storm_opendb: CONNECT error: Access denied for user 'storm'@'localhost' (using password: YES) 07/03 15:18:47 14106 storm_opendb: CONNECT error: Access denied for user 'storm'@'localhost' (using password: YES) 07/03 15:19:47 14106 storm_opendb: CONNECT error: Access denied for user 'storm'@'localhost' (using password: YES) 07/03 15:20:47 14106 storm_opendb: CONNECT error: Access denied for user 'storm'@'localhost' (using password: YES) 07/03 15:21:47 14106 storm_opendb: CONNECT error: Access denied for user 'storm'@'localhost' (using password: YES) 07/03 15:21:47 14106 storm_opendb: Too many connection attempts to the DB. Cowardly refusing to continue. 07/03 15:21:47 14106 srmv2: SRM02 - get_supported_protocols error : Cannot allocate memory
- Due to: config files are not modifyed according the user password set up in yaim config files
- Solution: modify password by hand in the following files.
/etc/sysconfig/srmv2storm.nsconfig -> /opt/srmv2storm/etc/sysconfig/srmv2storm.nsconfig /opt/srmv2storm/etc/db/storm_mysql_grant.sql
- Errors creating remote directories:
- logs showing errors like: "Caused by: Malformed SURL does not contains StFNRoot?: srm://ccc01.ific.uv.es:8444/atlas/XXX" .
- Connection errors using lcg_utils:
[alferca@cg02 srmtest]$ lcg-cr -v --vo dteam -d ccc01.ific.uv.es file:///tmp/kk.log Using grid catalog type: lfc Using grid catalog : lfc02.pic.es Using LFN : /grid/dteam/generated/2007-07-05/file-121bde62-fc39-4df1-a9e4-e215c403e72c Using SURL : srm://ccc01.ific.uv.es/lustre/ific.uv.es/dteam/generated/2007-07-05/filefc39f658-05d0-4c48-9b5f-6eac8ad24954 CGSI-gSOAP: Could not open connection ! lcg_cr: Connection refused
- The error seems to be related to the fact that it is not finding the endpoing for the SE ccc01 in the PPS BDII. Check the published info of the Storm server in the top BDII
- lcg_utils "The path specified in the SURL does not have a local equivalent":
- The following error is raised because of a wrong hadling of the srmv2.2 protocol by the GFAL high level tools (lcg-cr in this case). The tools does not create the proper directories on the remote SE node:
[alferca@cg02 srmtest]$ lcg-cr -v --vo dteam -d ccc01.ific.uv.es file:///tmp/kk.log Using grid catalog type: lfc Using grid catalog : lfc02.pic.es Using LFN : /grid/dteam/generated/2007-07-05/file-b1ecae0d-4958-4aa5-bf84-763d3e20fddc Using SURL : srm://ccc01.ific.uv.es/dteam/generated/2007-07-05/file5d81d810-4df8-4975-b8f8-c3b9fed42835 The path specified in the SURL does not have a local equivalent! lcg_cr: No such file or directory
- Versions of of GFAL and lcg_util tools < lcg_util-1.5.1-1 does not support srmv2.2 protocol.
- grid-ftp error when doing 3rd party transfers (as lcg-rep does): " the server sent an error response: 500 500 Illegal PORT Com"
[alferca@cg02 jdl]$ lcg-rep -v --vo atlas -d grid007g.cnaf.infn.it lfn:/grid/atlas/ific.uv.es/test.12208 Using grid catalog type: lfc Using grid catalog : lfc02.pic.es srm://ccc01.ific.uv.es/atlas/generated/2007-07-24/file175f15c4-da6a-470c-a8cd-597d2cbd4c88: srmPrepareToGet successfully handled! Source URL: lfn:/grid/atlas/ific.uv.es/test.12208 File size: 9 VO name: atlas Destination specified: grid007g.cnaf.infn.it Source URL for copy: gsiftp://ccc01.ific.uv.es:2811/lustre/ific.uv.es/atlas/generated/2007-07-24/file175f15c4-da6a-470c-a8cd-597d2cbd4c88 Destination URL for copy: gsiftp://grid007g.cnaf.infn.it/flatfiles/SE00/atlas/generated/2007-07-24/file4aa0dccf-f746-4628-bdd9-475c2043a896 # streams: 1 # set timeout to 0 0 bytes 0.00 KB/sec avg 0.00 KB/sec instthe server sent an error response: 500 500 Illegal PORT Command
- This is a configuration problem of the gridftp server at the storm node, that is not allowed by default to do passive transfers. Check the gridftp config .
- /etc/rc.d/init.d/storm-backend status does not work
[root@ccc01 root]# /etc/rc.d/init.d/storm-backend status
storm-backend: Usage: /etc/rc.d/init.d/storm-backend {start|stop|restart|force-reload|status|suspend|resume}
StoRM SRMv2 backend server [FAILED]
Workarounds and scripts
-- AlvaroFernandez - 1 Mar 2010 -- JavierSanchez - 04 Jun 2007- space-tokens-dynamicinfo_v2.tgz: space tokens dynamic info version2. Script to check dynamic info related to spaces at lustre and publish it in the info provider and update the storm mysql database
- test-srm.sh: battery of simple srm tests to check storm
Show attachments
Hide attachments
| I | Attachment | Action | Size | Date | Who | Comment |
|---|---|---|---|---|---|---|
 xml | namespace.xml | manage | 9.7 K | 23 Jul 2007 - 17:32 | AlvaroFernandez | old Namespace config file for storm |
 ldif | static-file-SE.ldif | manage | 6.6 K | 24 Sep 2007 - 09:50 | AlvaroFernandez | old gip ldif file of Storm in preproduction |
 tgz | space-tokens-dynamicinfo.tgz | manage | 10.0 K | 25 Nov 2008 - 15:01 | AlvaroFernandez | space tokens dynamic info |
| tgz | space-tokens-dynamicinfo_v2.tgz | manage | 10.0 K | 02 Mar 2010 - 11:36 | AlvaroFernandez | space tokens dynamic info version2. Script to check dynamic info related to spaces at lustre and publish it in the info provider and update the storm mysql database |
 sh | test-srm.sh | manage | 4.9 K | 02 Mar 2010 - 11:40 | AlvaroFernandez | battery of simple srm tests to check storm |
Atlas Web
Create New Topic
Index
Search
Changes
Notifications
Statistics
Preferences
Antares_KM3Net |
|
Copyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors. Ideas, requests, problems regarding TWiki? Send feedback