- Qlustar General - Qlustar Mailing Lists

[Qlustar-General] Qlustar Feature Update 02/20
by Roland Fehrenbacher 13 Feb '20

13 Feb '20

A number of new Qlustar features were completed recently and published together with the latest security updates ([1], [2]). QluMan 11.0.3.1 has the following new capabilities: * The 'Write Files' dialog [3] got a complete revamp: + The configuration files/links belonging to different QluMan 'Config Classes' are now neatly arranged in a single collapsible tree-view. + Colored LEDs have been introduced to indicate whether pending changes of the config files are waiting to be written and which files are affected. + It is now straightforward to cycle through the list of changed files exclusively, allowing for a fast overview of pending changes in diff format. + Selecting individual Config Classes to be written has become a lot simpler by providing checkboxes in the tree-view. + Overall, the new dialog greatly enhances the clarity of the cluster configuration space. It provides way better control and confidence when performing configuration changes. * Messages for config file writing have been added to the QluMan Log Viewer [4]. * The so-called QluMan 'save-button interface' was improved and implemented for the editing of command names in the Command Editor. Now name labels turn green when changed and black after being saved, improving transparency. If you use QluMan singularity images, make sure you download [5] the latest version. [1] https://www.qlustar.com/news/qsa-0210201-linux-kernel-vulnerabilities [2] https://www.qlustar.com/news/qsa-0210202-security-update-bundle [3] https://docs.qlustar.com/en-US/Qlustar_Cluster_OS/11.0/html-single/QluMan_G… [4] https://docs.qlustar.com/en-US/Qlustar_Cluster_OS/11.0/html-single/QluMan_G… [5] https://qlustar.com/download

1 0

[Qlustar-General] 2nd stage configuration failed at the first step.
by mohd.farid.ismail＠yandex.com 20 Jan '20

20 Jan '20

Hi, I'm having trouble getting pass the 1 boot process, which is to run /usr/sbin/qlustar-initial-config. After entering a name for the cluster, I always get the error HTTP access to the Qlustar repository is not possible Please check your curent network settings (received from DHCP): enp9s0: ..... Kernel IP routing table . . . Once you have made the required modifications, restart /usr/sbin/qlustar-initial-config! I can access the network (sshing to another machine), but cannot ping. It seems the DNS cannot be resolved (ping www.google.com results in the error "Temporary failure in name resolution.") Any hint on what I did wrong?

2 2

[Qlustar-General] Security/Bug fix releases 11.0.0.6-b515f1278 / 10.1.1.10-b509f1277
by Roland Fehrenbacher 09 Jan '20

09 Jan '20

The Qlustar releases 11.0.0.6-b515f1278 and 10.1.1.10-b509f1277 are ready including a number of security and bug fixes. Please check the following web pages for details about security fixes: https://qlustar.com/news/qsa-0109201-linux-kernel-vulnerabilities https://qlustar.com/news/qsa-0109202-security-update-bundle The following non-security related enhancements/bug fixes are included: * Qlustar 11.0.0.6-b515f1278 fixes a bug in disk setup configurations for netboot nodes using XFS on LVM

1 0

[Qlustar-General] upgrading qlustar from 10.1.1 to 11.0
by Tobias Moehle 08 Jan '20

08 Jan '20

Dear all, First of all I'd like to wish all developers/maintainers and users a happy new year and all the best. I know I am quite late with this topic, but other things distracted me so-far: I'd like to finally upgrade our cluster to 11.0. According to the upgrade instructions (https://docs.qlustar.com/en-US/Qlustar_Cluster_OS/11.0/html-single/Release_…) it is more or less an apt distupgrade; but since there are a lot of changes and I have very bad experience trying to upgrade merely ubuntu-versions (which usually breaks), I can't believe it is so simple? So I'd like to ask whether it is really as simple and what your experiences are? Any caveats, things to keep in mind? Can I do it while calculations are running or should I rather wait for all of them to finish (we have run-times of months, therefore this is a thing in my case). Any advice/comment is very welcome. Best regards, Tobias

3 3

[Qlustar-General] HDD requirements
by mohd.farid.ismail＠yandex.com 07 Jan '20

07 Jan '20

Hi, I was looking at the hardware requirement, and am wondering how strict is the hard disc 160GB or bigger in size requirement. What I mean is, lets say if I use a smaller than 160GB disc, will the installation stop itself, or will it install anyway but with problems later down the road? --farid

2 3

[Qlustar-General] Can't install
by Jarett DeAngelis 17 Dec '19

17 Dec '19

Hi, I have a small collection of Skylake HP desktops I'm deploying Qlustar onto as a lab environment. I cannot get the system to boot correctly. I'm using the Windows utility "Rufus" to image the USB stick. If I use "dd" mode, when I try to boot it skips right over it as though it weren't bootable. If I use "iso" mode, it gets about six seconds into the boot process, loading modules, before it complains it can't load the vfat module. It is indeed not in /lib! I feel like I must be doing something wrong here, but damned if I can figure it out solo. I have also tried this on a Mac and on a Linux machine with BalenaEtcher; same results as "iso" mode on Rufus. Hope everyone is having a good weekend! Jarett

2 5

[Qlustar-General] QluMan bugfix update 11.0.2.10
by Roland Fehrenbacher 17 Dec '19

17 Dec '19

A bug was discovered in the recently uploaded QluMan GUI version 11.0.2.8 that prevents the addition of new hosts. The bug is fixed in the most recent version 11.0.2.10. Please download the updated QluMan GUI singularity image 11.0.2 at https://qlustar.com/download that includes the fixed version.

1 0

[Qlustar-General] Re: Error: No such user: 'admin'
by Roland Fehrenbacher 13 Dec '19

13 Dec '19

>>>>> "TM" == Tobias Moehle <tobias.moehle(a)uni-rostock.de> writes: TM> Hi Roland, yes, with the newest version everything worked TM> without any problems. Great, thanks for the confirmation. Best, Roland

1 0

[Qlustar-General] Qlustar Feature Update 12/19
by Roland Fehrenbacher 09 Dec '19

09 Dec '19

A number of new Qlustar features were completed recently and published together with the latest security updates. QluMan has the following new capabilities: * RXEngine command output is searchable now [1], making the analysis of command results a lot easier. * RXEngine Command History management [2] has been improved to allow for fine-grained control during removal of unwanted entries. * Cluster logging as displayed in the QluMan Log Viewer [3] has been vastly enhanced: + New message categories have been added for managing chroots, start/stopping of slurmd and RXEngine executions in general. + Messages originating from node reboots have been merged into single summary entries to hugely reduce the number of messages and better track cluster state changes. * The QluMan GUI now checks for version updates when starting up [4]. If you use QluMan singularity images, make sure you download [5] the latest version. In addition, we have integrated the latest OpenHPC 1.3.9 release, so a big pile of updated packages are available for those of you running CentOS/OpenHPC. [1] https://docs.qlustar.com/en-US/Qlustar_Cluster_OS/11.0/html-single/QluMan_G… [2] https://docs.qlustar.com/en-US/Qlustar_Cluster_OS/11.0/html-single/QluMan_G… [3] https://docs.qlustar.com/en-US/Qlustar_Cluster_OS/11.0/html-single/QluMan_G… [4] https://docs.qlustar.com/en-US/Qlustar_Cluster_OS/11.0/html-single/QluMan_G… [5] https://qlustar.com/download

1 0

[Qlustar-General] Error: No such user: 'admin'
by Tobias Moehle 05 Dec '19

05 Dec '19

Dear all, I am trying to setup a new cluster using the current 11.0.0-3-image. I have tried already several times (also with updated image) and usually the setup works fine. However, when trying to create the token, I keep getting the error qluman-cli --gencert ERROR:client.cli.network:client.cli.network.Cluster.__init__(): could not connect to server Error: No such user: 'admin' ERROR:qlunet.Node:Channel[('zmq_version_info', 1)].do_recv(): exception in request generator Traceback (most recent call last): File "/usr/lib/python3/dist-packages/qluman-11/qluman-cli.py", line 1331, in gencert db.users.lookup(field="name", val=user) File "/usr/lib/python3/dist-packages/qluman-11/common/types.py", line 1866, in lookup raise KeyError KeyError During handling of the above exception, another exception occurred: Traceback (most recent call last): File "/usr/lib/python3/dist-packages/qlunet/Node.py", line 141, in do_recv args = next(self._generator) File "/usr/lib/python3/dist-packages/qluman-11/qluman-cli.py", line 1334, in gencert sys.exit(1) SystemExit: 1 ' /var/log/qlustar-initial-config.log ' does not contain any errors or suspiciously-looking parts and everything went through on the first trial. Thus I went out of own ideas about what could cause the problem. Many thanks in advance, Tobias Moehle

3 8

[Qlustar-General] failed to install lustre server
by Rajendra Adhikari 15 Nov '19

15 Nov '19

Dear All, Again greetings from Kathmandu University Supercomputer Center, Nepal. Please help me on installing the lustre server. Error message including my try are as below: I am not sure why there is dependency error. Sincerely, Rajendra Adhikari (mds) root@cl-head:~ # apt-cache search lustre | grep 'lustre-2.12-server-modules' lustre-2.12-server-modules-4.19.17-ql-generic-11.0-3 - Qlustar Lustre 2.12.x server kernel modules lustre-2.12-server-modules-4.19.31-ql-generic-11.0-4 - Qlustar Lustre 2.12.x server kernel modules lustre-2.12-server-modules-4.19.36-ql-generic-11.0-5 - Qlustar Lustre 2.12.x server kernel modules lustre-2.12-server-modules-4.19.42-ql-generic-11.0-6 - Qlustar Lustre 2.12.x server kernel modules lustre-2.12-server-modules-4.19.46-ql-generic-11.0-7 - Qlustar Lustre 2.12.x server kernel modules lustre-2.12-server-modules-ql-generic - Qlustar Lustre 2.12.x server kernel modules (Metapackage) lustre-2.12-server-modules-4.19.53-ql-generic-11.0-8 - Qlustar Lustre 2.12.x server kernel modules lustre-2.12-server-modules-4.19.60-ql-generic-11.0-9 - Qlustar Lustre 2.12.x server kernel modules lustre-2.12-server-modules-4.19.73-ql-generic-11.0-10 - Qlustar Lustre 2.12.x server kernel modules lustre-2.12-server-modules-4.19.80-ql-generic-11.0-11 - Qlustar Lustre 2.12.x server kernel modules I choose the last one: (mds) root@cl-head:~ # apt install lustre-2.12-server-modules-4.19.80-ql-generic-11.0-11 Reading package lists... Done Building dependency tree Reading state information... Done The following additional packages will be installed: lustre-2.12-common-modules-4.19.80-ql-generic-11.0-11 The following NEW packages will be installed: lustre-2.12-common-modules-4.19.80-ql-generic-11.0-11 lustre-2.12-server-modules-4.19.80-ql-generic-11.0-11 0 upgraded, 2 newly installed, 0 to remove and 0 not upgraded. 1 not fully installed or removed. Need to get 1,556 kB/2,884 kB of archives. After this operation, 20.3 MB of additional disk space will be used. Do you want to continue? [Y/n] Get:1 http://repo.qlustar.com/repo/ubuntu 11.0-bionic-proposed-updates/main amd64 lustre-2.12-common-modules-4.19.80-ql-generic-11.0-11 amd64 2.12.2+ds.1-ql.1+11.0-11 [1,556 kB] Fetched 1,556 kB in 2s (656 kB/s) debconf: delaying package configuration, since apt-utils is not installed (Reading database ... 20741 files and directories currently installed.) Preparing to unpack .../lustre-2.12-common-modules-4.19.80-ql-generic-11.0-11_2.12.2+ds.1-ql.1+11.0-11_amd64.deb ... Unpacking lustre-2.12-common-modules-4.19.80-ql-generic-11.0-11 (2.12.2+ds.1-ql.1+11.0-11) ... Preparing to unpack .../lustre-2.12-server-modules-4.19.80-ql-generic-11.0-11_2.12.2+ds.1-ql.1+11.0-11_amd64.deb ... Unpacking lustre-2.12-server-modules-4.19.80-ql-generic-11.0-11 (2.12.2+ds.1-ql.1+11.0-11) ... Setting up linux-image-4.19.80-ql-generic-11.0-11 (4.19-11.0-11) ... Running depmod. Failed to run depmod dpkg: error processing package linux-image-4.19.80-ql-generic-11.0-11 (--configure): installed linux-image-4.19.80-ql-generic-11.0-11 package post-installation script subprocess returned error exit status 1 dpkg: dependency problems prevent configuration of lustre-2.12-common-modules-4.19.80-ql-generic-11.0-11: lustre-2.12-common-modules-4.19.80-ql-generic-11.0-11 depends on linux-image-4.19.80-ql-generic-11.0-11; however: Package linux-image-4.19.80-ql-generic-11.0-11 is not configured yet. dpkg: error processing package lustre-2.12-common-modules-4.19.80-ql-generic-11.0-11 (--configure): dependency problems - leaving unconfigured dpkg: dependency problems prevent configuration of lustre-2.12-server-modules-4.19.80-ql-generic-11.0-11: lustre-2.12-server-modules-4.19.80-ql-generic-11.0-11 depends on linux-image-4.19.80-ql-generic-11.0-11; however: Package linux-image-4.19.80-ql-generic-11.0-11 is not configured yet. lustre-2.12-server-modules-4.19.80-ql-generic-11.0-11 depends on lustre-2.12-common-modules-4.19.80-ql-generic-11.0-11; however: Package lustre-2.12-common-modules-4.19.80-ql-generic-11.0-11 is not configured yet. dpkg: error processing package lustre-2.12-server-modules-4.19.80-ql-generic-11.0-11 (--configure): dependency problems - leaving unconfigured Errors were encountered while processing: linux-image-4.19.80-ql-generic-11.0-11 lustre-2.12-common-modules-4.19.80-ql-generic-11.0-11 lustre-2.12-server-modules-4.19.80-ql-generic-11.0-11 E: Sub-process /usr/bin/dpkg returned an error code (1) (mds) root@cl-head:~ # The output of uname -a is as follow: uname -a Linux cl-head 4.19.80-ql-generic-11.0-11 #1 SMP Mon Oct 21 09:49:44 UTC 2019 x86_64 x86_64 x86_64 GNU/Linux

2 3

[Qlustar-General] First time authentication
by Alain EMPAIN 13 Nov '19

13 Nov '19

Hello, I installed the new qlustar version 11.0.2 and tried to start qluman-qt : I cannot pass the first admin authentication step. 2019-11-13 11:35:17,207 DEBUG client.qt.ClusterWindow ClusterWindow.on_qlumand_online(True) 2019-11-13 11:35:17,208 DEBUG client.qt.ClusterWindow ClusterWindow.authenticate(): checking authentication 2019-11-13 11:35:17,208 DEBUG client.qt.ClusterWindow one time token 2019-11-13 11:35:17,208 DEBUG client.qt.MainWindow ServerWidget[b'Qlumand'].online_changed(True) 2019-11-13 11:35:17,213 DEBUG client.qt.ClusterWindow msg = error { text: "Authentication failed" } 2019-11-13 11:35:55,552 DEBUG client.qt.ClusterWindow one time token failed 2019-11-13 11:35:55,552 DEBUG client.qt.ClusterWindow ClusterWindow.closeEvent() 2019-11-13 11:35:55,576 DEBUG client.qt.ClusterWindow ClusterWindow.count_messages(force=True) 2019-11-13 11:35:55,576 DEBUG client.qt.ClusterWindow ClusterWindow._count_messages(force=True) 2019-11-13 11:35:55,576 DEBUG client.qt.ClusterWindow check_components(): components = [] 2019-11-13 11:35:55,576 DEBUG client.qt.MainWindow ServersWidget.cluster_activated(<client.qt.ClusterWindow.ClusterWindow object at 0x7fcdf8e4bfc8>) 2019-11-13 11:35:55,577 DEBUG client.qt.MainWindow no change 2019-11-13 11:35:55,582 DEBUG client.qt.ClusterWindow ClusterWindow.closeEvent(): all closed 2019-11-13 11:35:56,585 DEBUG client.qt.ClusterWindow ClusterWindow.closeEvent(): done I found the token files created at '/usr/lib/python3/dist-packages/qluman-11'. I erased them, to restart from a fresh state, but it does not help, I circle around the problem . When I installed the 10.x version, it worked immediately, but perhaps by chance ;-) Is it possible to reverse the 'first try procedure' to a sane state ? I suppose that my various attemps left some scories to be removed for a correct procedure. By the way, I copied the new 11.0.3 iso version on a USB key drive, with the classical command 'dd if=the.ISO of=/dev/sdb bs=...' but the server BIOS do not detect it as a bootable device. (I do this kind of operation routinely, nromally without any problem) Thanks for any advice, Alain --- Dr Alain EMPAIN, Bioinformatics University of Liège CHU, GIGA-R (B34, +1), alain.empain(a)uliege.be Rue des Martyrs, 11 B-4550 Nandrin Mobile: +32 497 701764 HOME:+32 85 512341

2 2

[Qlustar-General] qluman-cli --gencert error
by rajendra.adhikari＠ku.edu.np 11 Nov '19

11 Nov '19

Dear All, Please help me the problem with not able to generate qluman --gencert. The output is as below: Sincerely, Rajendra Adhikari qluman-cli --gencert ERROR:qlunet.Node:router is not responding to pings: reconnecting ERROR:qlunet.Node:Node.run(): exiting after exception Traceback (most recent call last): File "/usr/lib/python3/dist-packages/qlunet/Node.py", line 333, in run timeout = self._queue.next() File "/usr/lib/python3/dist-packages/qlunet/Timeout.py", line 143, in next item.expired(self) File "/usr/lib/python3/dist-packages/qlunet/Node.py", line 79, in expired self._node.slave_reset_keys() File "/usr/lib/python3/dist-packages/qlunet/Node.py", line 994, in slave_reset_keys self._monitor.setsockopt(zmq.LINGER, 0) AttributeError: 'NoneType' object has no attribute 'setsockopt' ERROR:qlunet.Node:Network thread is quiting on us

2 2

[Qlustar-General] Re: Problem to start qluman-qt : sqlite3 vs mysql
by Roland Fehrenbacher 05 Nov '19

05 Nov '19

>>>>> "A" == Alain EMPAIN <alain.empain(a)ulg.ac.be> writes: A> Hello Roland, thanks for the fast answer.. Indeed, just this A> command allows to start qlustar-qt Great, thanks for the confirmation. Fixed packages + singularity container are already uploaded. Please update your head-node as well, since another bug in the RXengine has been fixed. A> Have a nice day, Thanks, same to you. A> PS: I am very interested into Qlustar : it brings a fresh view on A> cluster management with respect to ROCKS. Before ROCKS, I was A> handling everything by hand (cluster itself, Condor, SGE, Torque A> engines...), so it was already a nice step towards ! Always great to hear positive statements. A> Here at the Liège University, and at the CECI consortium (Walloon A> Region), every clusters run SLURM and the official clusters are A> dealt with the commercial offer BRIGHT. I am building a fresh new A> cluster with Qlustar and old nodes, retired because we do not A> have maintenance contracts anymore for them ! I do not want to A> discard yet working machines, so I profit to experiment and to A> deploy Qlustar on a prototyping cluster, in the back of the deep A> basement of our Univ. Hospital ;-) Thanks for sharing your approach. I fully understand that one needs to follow a very conservative approach when trying to replace an existing working setup. Good luck with this. I'd be happy to hear more about your progress in the future here. Best, Roland A> From: "Roland Fehrenbacher" <rf(a)q-leap.de> To: "QLUSTAR A> General" <qlustar-general(a)qlustar.org> Sent: Tuesday, 5 A> November, 2019 09:46:15 Subject: [Qlustar-General] Re: A> Problem to start qluman-qt : sqlite3 vs mysql >>>>>> "A" == Alain EMPAIN <alain.empain(a)ulg.ac.be> writes: A> Hi Alain, A> Hello, I am discovering Qlustar (being a ROCKS admin since a long A> time). My first test with Qlustar 10.x was sucessful; qluman-qt A> started correctly, I browsed the windows with no problem. A> happy to see you trying Qlustar. A> I reinstalled everything from scratch to run Qlustar 11, A> installed qluman-qt with apt install and using the singularity A> image : both version give the error A> (...) File A> "/usr/lib/python3/dist-packages/sqlalchemy/engine/default.py", A> line 393, in connect return self.dbapi.connect(*cargs, **cparams) A> sqlalchemy.exc.OperationalError: (sqlite3.OperationalError) A> unable to open database file A> I wonder why SQLITE3 is invoqued because during the installation, A> there was a prompt for mysql root's password. A> The sqlite DB is a new feature we introduced with the latest A> updates to store user commands executed via the RXengine A> in. We assume that you discovered a small bug that we will A> fix shortly. As a work around you can try to execute A> $ mkdir $HOME/.config/Qluman A> This is where the sqlite DB should be stored and if the dir A> doesn't exist it will fail. Can you please let us know A> whether this fixes your problem? A> Best, A> Roland _______________________________________________ A> Qlustar-General mailing list -- qlustar-general(a)qlustar.org A> To unsubscribe send an email to A> qlustar-general-leave(a)qlustar.org --

1 0

[Qlustar-General] Problem to start qluman-qt : sqlite3 vs mysql
by Alain EMPAIN 05 Nov '19

05 Nov '19

Hello, I am discovering Qlustar (being a ROCKS admin since a long time). My first test with Qlustar 10.x was sucessful; qluman-qt started correctly, I browsed the windows with no problem. I reinstalled everything from scratch to run Qlustar 11, installed qluman-qt with apt install and using the singularity image : both version give the error (...) File "/usr/lib/python3/dist-packages/sqlalchemy/engine/default.py", line 393, in connect return self.dbapi.connect(*cargs, **cparams) sqlalchemy.exc.OperationalError: (sqlite3. OperationalError) unable to open database file I wonder why SQLITE3 is invoqued because during the installation, there was a prompt for mysql root's password. MariaDB is running, and a Qlustar Database exists and populated : mysql -p Enter password: Welcome to the MariaDB monitor. Commands end with ; or \g. Your MariaDB connection id is 48 Server version: 10.1.41-MariaDB-0ubuntu0.18.04.1 Ubuntu 18.04 Copyright (c) 2000, 2018, Oracle, MariaDB Corporation Ab and others. Type 'help;' or '\h' for help. Type '\c' to clear the current input statement. MariaDB [(none)]> use Qlustar; Reading table information for completion of table and column names You can turn off this feature to get a quicker startup with -A Database changed MariaDB [Qlustar]> select * from Users; +----+-------+--------+---------------+----------------+------------------------------------------------+---------------------+ | id | name | passwd | real_name | email | info | last_changed | +----+-------+--------+---------------+----------------+------------------------------------------------+---------------------+ | 1 | admin | $......| Admin account | root@localhost | Admin user | 2019-11-05 08:49:36 | | 2 | root | $......| cli user | root@beosrv-c | user used by the qluman command line interface | 2019-10-30 10:57:10 | +----+-------+--------+---------------+----------------+------------------------------------------------+---------------------+ 2 rows in set (0.00 sec) Any clue will be welcome ! Thanks Alain --- Dr Alain EMPAIN, Bioinformatics University of Liège CHU, GIGA-R (B34, +1), alain.empain(a)uliege.be Rue des Martyrs, 11 B-4550 Nandrin Mobile: +32 497 701764 HOME:+32 85 512341

2 1

[Qlustar-General] Feature Update 10/19 - QluMan RXEngine 2.0 / CentOS 7.7
by Roland Fehrenbacher 01 Nov '19

01 Nov '19

A large number of new exciting Qlustar features were completed recently and published as ** Feature Update 10/19 **. Most prominent is RXengine 2.0, a complete redesign of the QluMan Remote Execution engine. Non-QluMan related major new stuff includes the move to CentOS 7.7. Check out all the details at https://www.qlustar.com/news/qluman-rxengine-20-feature-update-1019

1 0

[Qlustar-General] Security/Bug fix releases 11.0.0.4-b515f1270 / 10.1.1.8-b509f1268
by Roland Fehrenbacher 30 Oct '19

30 Oct '19

The Qlustar releases 11.0.0.4-b515f1270 and 10.1.1.8-b509f1268 are ready including a number of security and bug fixes. Please check the following web pages for details about security fixes: https://qlustar.com/news/qsa-1024191-linux-kernel-vulnerabilities https://qlustar.com/news/qsa-1024192-security-update-bundle A number of non-security related enhancements were also part of this update and will be introduced in a separate announcement.

1 0

[Qlustar-General] run command after boot as specific user
by ignas.bunikis＠scilifelab.uu.se 15 Jul '19

15 Jul '19

Hej, What would be the best way to execute command after boot, as a specific user, only on VFend? Cheers, iggy

2 2

[Qlustar-General] how to use roles to limit rights for some admins
by Rolandas 01 Jul '19

01 Jul '19

Hi, I want to limit rights for admins in qluman. I tried to create operator role and define minimal rights. I created user in qluman with such role and after connecting with him, I found it could do all admin tasks, even those denied by role rules. What I'm doing wrong ? Regards Rolandas

3 3

[Qlustar-General] cannot run updated qluman gui
by Rolandas 27 Jun '19

27 Jun '19

Hi, ERROR : Unknown image format/type: qluman-10.1.1.23-singularity.sqsh ABORT : Retval = 255 singularity-container package is from ubuntu 19.04 repository: 2.6.1-2 Previous version qluman-10.1.1.19-singularity.sqsh works. MD5/SHA256 were verified and are the same as in qlustar dowanload page. Regards, Rolandas

3 6

[Qlustar-General] Support for Intel MIC Xeon Phi
by Roberto Rey Expósito 27 Jun '19

27 Jun '19

Hi, I wonder if it is possible to have Xeon Phi hardware on a Qlustar-based cluster. I cannot see any module related with Xeon Phi during head node installation as it is the case for GPU or InfiniBand hardware, so the resulting CentOS image for compute nodes does not recognize the cards. Thanks, -- Roberto R. Expósito Grupo de Arquitectura de Computadores (GAC) Depto. Ingeniería de Computadores Universidade da Coruña (UDC)

2 3

[Qlustar-General] Qlustar server in HA mode
by darkhalf78＠gmail.com 25 Jun '19

25 Jun '19

Hello, sorry if the question has been already discussed in previous threads, but i was wondering if there is a guide on how to configure two physical servers in HA mode as stated in the Administration Manual "In case of a HA head-node setup, the second head-node becomes a NIS slave server. " I configured the second server as NIS slave, but i'm more interested into a redundant configuration with drbd for the whole server. I'm trying to configure drbd to mirror all the partitions/LV on the second server using LCMC but with no success. Any guide or hint is very welcome :) Thanks a lot !

2 1

[Qlustar-General] Qlustar 11 is out
by Roland Fehrenbacher 14 Jun '19

14 Jun '19

Check out the release notes at https://docs.qlustar.com/en-US/Qlustar_Cluster_OS/11.0/html-single/Release_… Enjoy, Roland

1 0

[Qlustar-General] cannot start MPI jobs
by Rolandas 13 Jun '19

13 Jun '19

Hi, In fresh qlustar installation without infiniband we cannot start MPI jobs. Error message is long, but essential part is: [beo-01:19012] pmix_mca_base_component_repository_open: "mca_bfrops_v21" does not appear to be a valid bfrops MCA dynamic component (ignored): /usr/lib/openmpi/3.1.2/gcc/lib/openmpi/mca_pmix_ext2x.so: undefined symbol: mca_bfrops_v21_component. ret -1 Qlustar image for nodes have modules: core, slurm, ofed, lustre-client. UnionFS Chroot (xenial/10.1) is freshly installed without additional packages installed (dist-upgrade was executed). Regards, Rolandas

2 2

[Qlustar-General] Slurm srun & MPI
by Ansgar Esztermann-Kirchner 12 Jun '19

12 Jun '19

Hello List, according to https://slurm.schedmd.com/mpi_guide.html#intel_srun, srun should be able to lauch MPI jobs on its own. However, my test jobs fail with errors like these: Fatal error in PMPI_Init_thread: Other MPI error, error stack: MPIR_Init_thread(805).....: fail failed MPID_Init(1832)...........: channel initialization failed ... Prior to the above errors, I see: mdrun_mpi_AVX_256: /usr/lib/x86_64-linux-gnu/slurm/auth_munge.so: Incompatible Slurm plugin version (17.11.9) mdrun_mpi_AVX_256: /usr/lib/x86_64-linux-gnu/slurm/auth_munge.so: Incompatible Slurm plugin version (17.11.9) mdrun_mpi_AVX_256: error: Couldn't load specified plugin name for auth/munge: Incompatible plugin version mdrun_mpi_AVX_256: error: cannot create auth context for auth/munge Where mdrun_mpi_AVX_256 is my executable. I've regenerated qlustar images and rebooted the nodes, but to no avail. I've also checked that md5sums for auth_munge.so and slurmstepd match the ones given in /usr/lib/qlustar/modules/xenial-amd64/10.1.1/slurm.contents, and now I'm out of ideas. I gues it would be nice to know what slurm thinks the correct version is, but unfortunately, only the offending version is printed... Any ideas what I should try next? Thanks, A. -- Ansgar Esztermann Sysadmin Dep. Theoretical and Computational Biophysics http://www.mpibpc.mpg.de/grubmueller/esztermann

2 4

2025

2024

2023

2022

2021

2020

2019

2018

Qlustar General