We generally recommend for pkglist to be "the list of packages shipped in Linux distribution ISO which will be installed on the node." And for otherpkglist to be "list of user specified additional packages not shipped in Linux distribution ISO which will be installed on the node."
Are the packages in the "warning" list listed in otherpkglist and located in the otherpkgdir ? Mark Gurevich Poughkeepsie Development Lab HPC Software Development - xCAT "If we knew what it was we were doing, it would not be called research, would it?" --Albert Einstein From: "Huette, Antoine" <antoine.hue...@bechtle.com> To: xCAT Users Mailing list <xcat-user@lists.sourceforge.net> Date: 09/24/2020 07:47 AM Subject: [EXTERNAL] Re: [xcat-user] repomd.xml file Hi, So, after recreating the pkglist and moving some packages from pkglist to other.pkglist, the packages can now be sent to compute nodes. However, there’s now an error while running the otherpkgs postbootscript. The error is : Warning: the packages munge-libs* munge* perl-Switch* slurm* slurm-contribs* slurm-devel* slurm-example-configs* slurm-libpmi* slurm-pam_slurm* slurm-perlapi* slurm-slurmd* ganglia* ganglia-devel* ganglia-gmond* ganglia-gmond-python* libconfuse* could not be found in the yum/apt repository, falling back to rpm/dpkg command. If you want your packages to be installed with yum/apt, verify yum/apt is installed and createrepo/dpkg-scanpackages has been run. For some reason it seems compute nodes try to get other packages from online repos, because then there’s another error saying it couldn’t find a valid baseurl for repo base/7/x86_64 and it couldn’t retrieve mirrorlist from mirrorlist.centos.org How do I prevent compute nodes in the installation process to get packages from other sources than the mgmt node ? I don’t remember having to do such a thing in the past… Regards, Antoine Huette Advisor Bechtle direct SAS Le Xenium - Rue des Vergers 30, FR-67120 Molsheim Phone: 33367079737 E-Mail: antoine.hue...@bechtle.com Video: JabberGuest Web: bechtle.fr DSGVO/GDPR: Privacy Policy Art. 13,14 De : Huette, Antoine Envoyé : jeudi 24 septembre 2020 09:04 À : xCAT Users Mailing list <xcat-user@lists.sourceforge.net> Objet : RE: [xcat-user] repomd.xml file Hi, Not much : #Please make sure there is a space between @ and group name wget ntp nfs-utils net-snmp rsync yp-tools openssh-server util-linux net-tools gcc gcc-c++ gcc-gfortran yum-utils mlocate bind-utils bash bash-completion vim nano sed gawk emacs createrepo kernel-devel xorg-x11-server-Xorg xorg-x11-xauth ypbind rpcbind yp-tools munge ganglia-devel ganglia-gmond ganglia-gmond-python ganglia-web ganglia openldap-clients nss-pam-ldapd But when I try with the default pkglist file, the same issue occurs. I've checked all packages listed in the default file exist in the os repo, and they are all located here. Regards, Antoine Huette Advisor Bechtle direct SAS Le Xenium - Rue des Vergers 30, FR-67120 Molsheim Phone: 33367079737 E-Mail: antoine.hue...@bechtle.com Video: JabberGuest Web: bechtle.fr DSGVO/GDPR: Privacy Policy Art. 13,14 De : Mark Gurevich <gurev...@us.ibm.com> Envoyé : mercredi 23 septembre 2020 19:38:48 À : xCAT Users Mailing list Objet : Re: [xcat-user] repomd.xml file And what is listed in your packagelist /opt/xcat/share/xcat/install/centos/compute.centos7.pkglist ? Mark Gurevich Poughkeepsie Development Lab HPC Software Development - xCAT "If we knew what it was we were doing, it would not be called research, would it?" --Albert Einstein Inactive hide details for "Huette, Antoine" ---09/23/2020 11:24:14 AM---Hi, I realized it uses wget because when I remove the p"Huette, Antoine" ---09/23/2020 11:24:14 AM---Hi, I realized it uses wget because when I remove the proxy conf in the wgetrc file, the xcatprobe c From: "Huette, Antoine" <antoine.hue...@bechtle.com> To: xCAT Users Mailing list <xcat-user@lists.sourceforge.net> Date: 09/23/2020 11:24 AM Subject: [EXTERNAL] Re: [xcat-user] repomd.xml file Hi, I realized it uses wget because when I remove the proxy conf in the wgetrc file, the xcatprobe command does not report a single error now :) When I do a wget on a RPM file from the mgmt or another server it works just fine. It looks like there's something that prevents compute nodes to know they have to retrieve the RPMs... Best regards, Antoine Huette Advisor Bechtle direct SAS Le Xenium - Rue des Vergers 30, FR-67120 Molsheim Phone: 33367079737 E-Mail: antoine.hue...@bechtle.com Video: JabberGuest Web: bechtle.fr DSGVO/GDPR: Privacy Policy Art. 13,14 De : Mark Gurevich <gurev...@us.ibm.com> Envoyé : mercredi 23 septembre 2020 16:32:44 À : xCAT Users Mailing list Objet : Re: [xcat-user] repomd.xml file This HTTP service failure reported by "xcatprobe" could be a clue. To generate that error "xcatprobe" tries to "wget" a file from "http://<MN IP>:<httpport>/<installdir>/postscripts/" The "httpport" and "installdir" are discovered by running "lsdef -t site -i installdir,httpport -c", if httpport is not specified, default is 80. You can try to manually on MN wget a file from that directory and see if any errors are reported. Mark Gurevich Poughkeepsie Development Lab HPC Software Development - xCAT "If we knew what it was we were doing, it would not be called research, would it?" --Albert Einstein "Huette, Antoine" ---09/23/2020 08:20:25 AM---Just in case, the httpd access.log file contains the following : 192.168.1.117 - - [23/Sep/2020:14:0 From: "Huette, Antoine" <antoine.hue...@bechtle.com> To: xCAT Users Mailing list <xcat-user@lists.sourceforge.net> Date: 09/23/2020 08:20 AM Subject: [EXTERNAL] Re: [xcat-user] repomd.xml file Just in case, the httpd access.log file contains the following : 192.168.1.117 - - [23/Sep/2020:14:07:39 +0200] "GET /tftpboot/xcat/xnba/nodes/node17 HTTP/1.1" 200 555 "-" "iPXE/1.0.3-131028 (d603e)" 192.168.1.117 - - [23/Sep/2020:14:07:39 +0200] "GET /tftpboot/xcat/osimage/centos7.7-x86_64-install-compute/vmlinuz HTTP/1.1" 200 6734016 "-" "iPXE/1.0.3-131028 (d603e)" 192.168.1.117 - - [23/Sep/2020:14:07:39 +0200] "GET /tftpboot/xcat/osimage/centos7.7-x86_64-install-compute/initrd.img HTTP/1.1" 200 55073584 "-" "iPXE/1.0.3-131028 (d603e)" 192.168.1.117 - - [23/Sep/2020:14:08:08 +0200] "GET /install/centos7.7/x86_64/.treeinfo HTTP/1.1" 200 354 "-" "curl/7.29.0" 192.168.1.117 - - [23/Sep/2020:14:08:08 +0200] "GET /install/centos7.7/x86_64/LiveOS/squashfs.img HTTP/1.1" 200 498380800 "-" "curl/7.29.0" 192.168.1.117 - - [23/Sep/2020:14:08:12 +0200] "GET /install/centos7.7/x86_64/images/updates.img HTTP/1.1" 404 241 "-" "curl/7.29.0" 192.168.1.117 - - [23/Sep/2020:14:08:13 +0200] "GET /install/centos7.7/x86_64/images/product.img HTTP/1.1" 404 241 "-" "curl/7.29.0" 192.168.1.117 - - [23/Sep/2020:14:08:13 +0200] "GET /install/autoinst/node17 HTTP/1.1" 200 98095 "-" "curl/7.29.0" 192.168.1.117 - - [23/Sep/2020:14:08:29 +0200] "GET /install/centos7.7/x86_64/.treeinfo HTTP/1.1" 200 354 "-" "urlgrabber/3.10" 192.168.1.117 - - [23/Sep/2020:14:08:29 +0200] "GET /install/centos7.7/x86_64/repodata/repomd.xml HTTP/1.1" 200 3733 "-" "CentOS (anaconda)/7 yum/3.4.3" 192.168.1.117 - - [23/Sep/2020:14:08:29 +0200] "GET /install/centos7.7/x86_64/.treeinfo HTTP/1.1" 200 354 "-" "urlgrabber/3.10" 192.168.1.117 - - [23/Sep/2020:14:08:29 +0200] "GET /install/centos7.7/x86_64/repodata/03dde1157310b4a6d8e9128de19651c147a71150d406f3efaaf12f7eb83a1f35-primary.xml.gz HTTP/1.1" 200 1614194 "-" "CentOS (anaconda)/7 yum/3.4.3" 192.168.1.117 - - [23/Sep/2020:14:08:30 +0200] "GET /install/centos7.7/x86_64/repodata/4af1fba0c1d6175b7e3c862b4bddfef93fffb84c37f7d5f18cfbff08abc47f8a-c7-x86_64-comps.xml.gz HTTP/1.1" 200 169182 "-" "CentOS (anaconda)/7 yum/3.4.3" 192.168.1.117 - - [23/Sep/2020:14:08:30 +0200] "GET /install/centos7.7/x86_64/repodata/7a2197223a2689b44594b09351dead84fe2d8cd0459be3f885bb985bb28af7ca-primary.sqlite.bz2 HTTP/1.1" 200 3389624 "-" "CentOS (anaconda)/7 yum/3.4.3" 192.168.1.117 - - [23/Sep/2020:14:08:30 +0200] "GET /install/centos7.7/x86_64//repodata/repomd.xml HTTP/1.1" 200 3733 "-" "urlgrabber/3.10" I've also regenerated the os directory with the copycds command (tried with both DVD and Everything isos), just to be sure nothing wrong happened the first time, but no change. Best regards Antoine Huette Advisor Bechtle direct SAS Le Xenium - Rue des Vergers 30, FR-67120 Molsheim Phone: 33367079737 E-Mail: antoine.hue...@bechtle.com Video: JabberGuest Web: bechtle.fr DSGVO/GDPR: Privacy Policy Art. 13,14 De : Huette, Antoine Envoyé : mercredi 23 septembre 2020 09:24:50 À : xCAT Users Mailing list Objet : RE: [xcat-user] repomd.xml file Hi, Yes, the RPMs have read permissions. In computes.log and cluster.log the last message is "Generate de repository for the installation" xcatprobe xcatmn says the following : [mn]: Checking HTTP service is configured... [FAIL] [mn]: The server, while acting as a gateway or proxy, did not receive a timely response from the upstream serve... Could it be related ? The httpd service seems to be fine though... Best regards, Antoine Huette Advisor Bechtle direct SAS Le Xenium - Rue des Vergers 30, FR-67120 Molsheim Phone: 33367079737 E-Mail: antoine.hue...@bechtle.com Video: JabberGuest Web: bechtle.fr DSGVO/GDPR: Privacy Policy Art. 13,14 De : Mark Gurevich <gurev...@us.ibm.com> Envoyé : mardi 22 septembre 2020 20:01:07 À : xCAT Users Mailing list Objet : Re: [xcat-user] repomd.xml file After receiving the repomd.xml file, the compute node should start receiving the packages from directory specified by pkgdir. Check that the rpms there have read permissions. You can also check /var/log/xcat/computes.log, /var/log/xcat/cluster.log and /var/log/messages to see if any errors reported. When I provision one of my nodes with RH7.7, xcatprobe displays: [c910f03c09k18] 13:47:20 INFO ============deployment starting============ [c910f03c09k18] 13:47:20 INFO Running Anaconda Pre-Installation script... [c910f03c09k18] 13:47:20 INFO Running Anaconda Pre-Installation script... [c910f03c09k18] 13:47:20 INFO Detecting install disk... [c910f03c09k18] 13:47:20 INFO Detecting install disk... [c910f03c09k18] 13:47:20 Node status is changed to installing [c910f03c09k18] 13:47:22 INFO Found /dev/sda, generate partition file... [c910f03c09k18] 13:47:22 INFO Found /dev/sda, generate partition file... [c910f03c09k18] 13:47:22 INFO Generate the repository for the installation [c910f03c09k18] 13:47:22 INFO Generate the repository for the installation [c910f03c09k18] 13:47:27 Via HTTP get /install/rhels7.7/ppc64le/.treeinfo [c910f03c09k18] 13:47:27 Via HTTP get /install/rhels7.7/ppc64le/repodata/repomd.xml [c910f03c09k18] 13:47:27 Via HTTP get /install/rhels7.7/ppc64le/.treeinfo [c910f03c09k18] 13:47:27 Via HTTP get /install/rhels7.7/ppc64le/repodata/ad215b83530ed774c937101631fc11f4eeb18... [c910f03c09k18] 13:47:27 Via HTTP get /install/rhels7.7/ppc64le/repodata/38fb729a35a94b2327e558e84dbc07efdc06e... [c910f03c09k18] 13:47:29 Via HTTP get /install/rhels7.7/ppc64le//repodata/repomd.xml [c910f03c09k18] 13:47:43 Via HTTP get /install/rhels7.7/ppc64le/Packages/perl-HTTP-Tiny-0.033-3.el7.noarch.rpm [c910f03c09k18] 13:47:43 Via HTTP get /install/rhels7.7/ppc64le/Packages/libthai-0.1.14-9.el7.ppc64le.rpm [c910f03c09k18] 13:47:43 Via HTTP get /install/rhels7.7/ppc64le/Packages/systemd-219-67.el7.ppc64le.rpm [c910f03c09k18] 13:47:43 Via HTTP get /install/rhels7.7/ppc64le/Packages/qrencode-libs-3.4.1-3.el7.ppc64le.rpm : : : Mark Gurevich Poughkeepsie Development Lab HPC Software Development - xCAT "If we knew what it was we were doing, it would not be called research, would it?" --Albert Einstein Inactive hide details for "Huette, Antoine" ---09/22/2020 11:32:12 AM---Hi, Yes sure, here is the output of lsdef (bmc and mac "Huette, Antoine" ---09/22/2020 11:32:12 AM---Hi, Yes sure, here is the output of lsdef (bmc and mac removed for security reasons, but they're goo From: "Huette, Antoine" <antoine.hue...@bechtle.com> To: xCAT Users Mailing list <xcat-user@lists.sourceforge.net> Date: 09/22/2020 11:32 AM Subject: [EXTERNAL] Re: [xcat-user] repomd.xml file Hi, Yes sure, here is the output of lsdef (bmc and mac removed for security reasons, but they're good) : Object name: node17 arch=x86_64 bmc=192.168.3.17 bmcpassword=<...> bmcusername=<...> currchain=boot currstate=install centos7.7-x86_64-compute groups=compute,all ip=192.168.1.17 mac=<...> mgt=ipmi netboot=xnba nicips.ib0=192.168.2.17 nicnetworks.ib0=192_168_2_0-255_255_255_0 nictypes.ib0=Infiniband os=centos7.7 postbootscripts=otherpkgs,confignics postscripts=syslog,remoteshell,syncfiles,mlnxofed_ib_install -p /install/MLNX_OFED_LINUX-5.1-2.3.7.1-rhel7.7-x86_64.iso,custom profile=compute provmethod=centos7.7-x86_64-install-compute serialport=0 serialspeed=115200 status=installing statustime=09-22-2020 15:42:59 Object name: centos7.7-x86_64-install-compute imagetype=linux osarch=x86_64 osdistroname=centos7.7-x86_64 osname=Linux osvers=centos7.7 otherpkgdir=/install/post/otherpkgs/centos7.7/x86_64 otherpkglist=/opt/xcat/share/xcat/install/centos/compute.centos7.other.pkglist pkgdir=/install/centos7.7/x86_64 pkglist=/opt/xcat/share/xcat/install/centos/compute.centos7.pkglist profile=compute provmethod=install template=/opt/xcat/share/xcat/install/centos/compute.centos7.tmpl And here is what xcatprobe says : The install NIC in current server is eno2 [INFO] All nodes to be deployed are valid [ OK ] ------------------------------------------------------------- Start capturing every message during OS provision process.... ------------------------------------------------------------- [node17] 14:57:54 Via TFTP download xcat/xnba.kpxe [node17] 14:57:54 Via TFTP download xcat/xnba.kpxe [node17] 14:57:55 Via HTTP get /tftpboot/xcat/xnba/nodes/node17 [node17] 14:57:55 Via HTTP get /tftpboot/xcat/osimage/centos7.7-x86_64-install-compute/vmlinuz [node17] 14:57:55 Via HTTP get /tftpboot/xcat/osimage/centos7.7-x86_64-install-compute/initrd.img [node17] 14:58:23 Via HTTP get /install/centos7.7/x86_64/.treeinfo [node17] 14:58:23 Via HTTP get /install/centos7.7/x86_64/LiveOS/squashfs.img [node17] 14:58:27 Via HTTP get /install/centos7.7/x86_64/images/updates.img [node17] 14:58:27 Via HTTP get /install/centos7.7/x86_64/images/product.img [node17] 14:58:28 Via HTTP get /install/autoinst/node17 [node17] 14:58:41 INFO ============deployment starting============ [node17] 14:58:41 INFO ============deployment starting============ [node17] 14:58:41 INFO Running Anaconda Pre-Installation script... [node17] 14:58:41 INFO Running Anaconda Pre-Installation script... [node17] 14:58:41 INFO Detecting install disk... [node17] 14:58:41 INFO Detecting install disk... [node17] 14:58:42 Node status is changed to installing [node17] 14:58:42 INFO Found /dev/sda, generate partition file... [node17] 14:58:42 INFO Found /dev/sda, generate partition file... [node17] 14:58:42 INFO Generate the repository for the installation [node17] 14:58:42 INFO Generate the repository for the installation [node17] 14:58:44 Via HTTP get /install/centos7.7/x86_64/.treeinfo [node17] 14:58:45 Via HTTP get /install/centos7.7/x86_64/repodata/repomd.xml [node17] 14:58:45 Via HTTP get /install/centos7.7/x86_64/.treeinfo [node17] 14:58:45 Via HTTP get /install/centos7.7/x86_64/repodata/afe9c5f88d9c365382c0553424b18192b79f3da7c1e3d... [node17] 14:58:45 Via HTTP get /install/centos7.7/x86_64/repodata/4af1fba0c1d6175b7e3c862b4bddfef93fffb84c37f7d... [node17] 14:58:45 Via HTTP get /install/centos7.7/x86_64/repodata/04efe80d41ea3d94d36294f7107709d1c8f70db11e152... [node17] 14:58:45 Via HTTP get /install/centos7.7/x86_64//repodata/repomd.xml Regards, Antoine Huette Advisor Bechtle direct SAS Le Xenium - Rue des Vergers 30, FR-67120 Molsheim Phone: 33367079737 E-Mail: antoine.hue...@bechtle.com Video: JabberGuest Web: bechtle.fr DSGVO/GDPR: Privacy Policy Art. 13,14 De : Mark Gurevich <gurev...@us.ibm.com> Envoyé : mardi 22 septembre 2020 16:56:28 À : xCAT Users Mailing list Objet : Re: [xcat-user] repomd.xml file Can you show the output of "lsdef" of the node and the osimage you are deploying ? Also, can you show the output of the "xcatprobe osdeploy" Mark Gurevich Poughkeepsie Development Lab HPC Software Development - xCAT "If we knew what it was we were doing, it would not be called research, would it?" --Albert Einstein Inactive hide details for "Huette, Antoine" ---09/22/2020 10:35:28 AM---Hi On a CentOS 7.7 cluster with latest xcat, compute n"Huette, Antoine" ---09/22/2020 10:35:28 AM---Hi On a CentOS 7.7 cluster with latest xcat, compute nodes cannot be deployed. There is an issue wh From: "Huette, Antoine" <antoine.hue...@bechtle.com> To: xCAT Users list <xcat-user@lists.sourceforge.net> Date: 09/22/2020 10:35 AM Subject: [EXTERNAL] [xcat-user] repomd.xml file Hi On a CentOS 7.7 cluster with latest xcat, compute nodes cannot be deployed. There is an issue while sending the /install/CentOs7.7/x86_64/repodata/repomd.xml file, or at least this is the latest message appearing in the xcatprobe osdeploy tool. I’ve tried regenerating osimage and the repository metadata but that did nothing. What’s the problem here ? I’m not sure where to look at. Regards Antoine Huette HPC engineer Bechtle _______________________________________________ xCAT-user mailing list xCAT-user@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/xcat-user _______________________________________________ xCAT-user mailing list xCAT-user@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/xcat-user _______________________________________________ xCAT-user mailing list xCAT-user@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/xcat-user _______________________________________________ xCAT-user mailing list xCAT-user@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/xcat-user _______________________________________________ xCAT-user mailing list xCAT-user@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/xcat-user
_______________________________________________ xCAT-user mailing list xCAT-user@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/xcat-user
