February 2022 - Qlustar General - Qlustar Mailing Lists

users created with LDAP in QluMan can't connect to head node with ssh
by saeed.s.jahromi＠gmail.com 22 Feb '22

22 Feb '22

Greetings I have installed the qlustar on a small cluster with one head node and 15 compute nodes. I managed to add all nodes to the cluster and configure slurm on the cluster via QluMan. As the last step to make the cluster available to users, I started creating new groups and users with QluMan. after creating new users and syncing I tried to connect with users to the head node by ssh. However, all the users that I create can not connect to the head node because it says the user does not have premission. I also chose to create a virtual demo cluster during the installation and the demo user which is created during the installation also does not have permission to ssh login to the head node. I checked on the head node and all user's home directory is created in the /data/home/ directory but they can not connect to the cluster. I would be so thankful if you guide me on how should I resolve this issue or where in the qlustar config files should I look for the source of the problem. PS, this cluster was already using an older version of the qlustar from 6 years ago and we just updated it to the latest qlustar 12. Thanks in advance for your assistance. Saeed

3 5

AlmaLinux 8 substitutes CentOS 8
by Qlustar List 21 Feb '22

21 Feb '22

https://qlustar.com/news/2022/2022-02-21-almalinux-8

1 0

How to generate a image for tesla m2075 ?
by Gerhard W. Recher 21 Feb '22

21 Feb '22

Dear members, I just test transition from rocks cluster to qlustar. one point is our tesla m2075 nodes wont boot with nvidia driver due to not supportet by ubuntou-focal-nvidea. <snip> [ 60.417821] NVRM: The NVIDIA Tesla M2075 GPU installed in this system is NVRM: supported through the NVIDIA 390.xx Legacy drivers. Please NVRM: visit http://www.nvidia.com/object/unix.html for more NVRM: information. The 470.74 NVIDIA driver will ignore NVRM: this GPU. Continuing probe... [ 60.418062] NVRM: No NVIDIA GPU found. [ 60.418406] nvidia-nvlink: Unregistered the Nvlink Core, major device number 245 [ 65.107363] nvidia-nvlink: Nvlink Core is being initialized, major device number 245 <snap> so how can i achieve this on head-node to assemble a new boot image with older driver ? fail with this one: NVIDIA-Linux-x86_64-352.79.run nvidia-installer log file '/var/log/nvidia-installer.log' creation time: Sun Feb 20 13:33:54 2022 installer version: 352.79 PATH: /usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/games:/usr/local/games:/snap/bin nvidia-installer command line: ./nvidia-installer Unable to load: nvidia-installer ncurses user interface Using built-in stream user interface -> Detected 4 CPUs online; setting concurrency level to 4. WARNING: You do not appear to have an NVIDIA GPU supported by the 352.79 NVIDIA Linux graphics driver installed in this system. For further details, please see the appendix SUPPORTED NVIDIA GRAPHICS CHIPS in the README available on the Linux driver download page at www.nvidia.com. -> License accepted. -> Installing NVIDIA driver version 352.79. -> Performing CC sanity check with CC="/usr/bin/cc". -> Kernel source path: '/lib/modules/5.4.167-ql-generic-12.0-13/build' -> Kernel output path: '/lib/modules/5.4.167-ql-generic-12.0-13/build' -> Performing rivafb check. -> Performing nvidiafb check. -> Performing Xen check. -> Performing PREEMPT_RT check. -> Cleaning kernel module build directory. executing: 'cd ./kernel; /usr/bin/make clean'... -> Building NVIDIA kernel module: executing: 'cd ./kernel; /usr/bin/make module SYSSRC=/lib/modules/5.4.167-ql-generic-12.0-13/build SYSOUT=/lib/modules/5.4.167-ql-generic-12.0-13/build -j4 NV_BUILD_MODULE_INSTANCES='... NVIDIA: calling KBUILD... make[1]: Entering directory '/usr/src/linux-headers-5.4.167-ql-generic-12.0-13' /usr/bin/make -f ./Makefile syncconfig /usr/bin/make -f ./scripts/Makefile.build obj=scripts/basic rm -f .tmp_quiet_recordmcount /usr/bin/make -f ./scripts/Makefile.build obj=scripts/kconfig syncconfig scripts/kconfig/conf --syncconfig Kconfig Kconfig:34: can't open file "Documentation/Kconfig" make[3]: *** [scripts/kconfig/Makefile:73: syncconfig] Error 1 make[2]: *** [Makefile:590: syncconfig] Error 2 make[1]: *** [Makefile:696: include/config/auto.conf.cmd] Error 2 make[1]: Leaving directory '/usr/src/linux-headers-5.4.167-ql-generic-12.0-13' NVIDIA: left KBUILD. nvidia.ko failed to build! make: *** [nvidia-modules-common.mk:245: module] Error 1 -> Error. ERROR: Unable to build the NVIDIA kernel module. ERROR: Installation has failed. Please see the file '/var/log/nvidia-installer.log' for details. You may find suggestions on fixing installation problems in the README available on the Linux driver download page at www.nvidia.com. -- Gerhard W. Recher net4sec UG (haftungsbeschränkt) Leitenweg 6 86929 Penzing +49 8191 4283888 +49 171 4802507

1 0

tcsh segfault in Ubuntu 20.04 with external LDAP
by a.gabbana＠tue.nl 17 Feb '22

17 Feb '22

Dear Roland, we observe the following reproducible problem: - A fresh installation of qlustar 12 - A chroot with Ubuntu 20.04 - An external (fresh) ldap server (connection via shared conf files under /etc/qlustar/common/image-files) Calling tcsh returns segfault. The problem seems to be connected to libgssapi_krb5.so.2. The problem does not occur with a Ubuntu 18.04 chroot. Moreover, the problem does not occur with a fresh installation of Ubuntu 20.04 (without qlustar) connected to the same ldap Do you have any clue what could be triggering the problem / or can you suggest a possible workaround? best Alessandro

2 1