I'd like to run htop on a few hosts, using clusterssh. I haven't changed any defaults in .clusterssh/config. This works fine: $ clusterssh somehosts -a uptime ... 12:11:30 up 8 days, 4:42, 1 user, load average: 2.04, 1.74, 1.71 This doesn't: $ clusterssh somehosts -a htop ... Error opening terminal: unknown. If I run without an action, TERM is set: $ clusterssh somehosts ... $ echo $TERM xterm Can anyone point me toward useful info on this? TIA.

I am using cssh on Debian 10 (package clusterssh 4.13.2-2). when I connect to multiple servers: cssh foo bar zz aa cssh will open 4 terminals and **ORDERS THEM ALPHABETICALLY !!!** So that I get the windows ordered on my screen as follows: aa bar foo zz This is pure evil ! How can I keep the windows in original order? Is there some quick fix in the perl code?

I want to submit a task that is interpreted by /bin/csh, which only exists in master node. And I have no root permission but only sudo, which is limited in master node. So I can't use sudo apt install csh in each calculation node. How to handle the case?

To monitor the job status in clusters,

is used to output lines like this

job-ID  prior   name       user         state submit/start at     queue                          slots ja-task-ID
-----------------------------------------------------------------------------------------------------------------
 146767 2.75000 REMD       xxxxxx      Rr    03/26/2021 10:58:17 Arya@node-c11b-027.kathleen.uc   160
 146811 2.75000 REMD       xxxxxx      r     03/26/2021 11:37:48 Arya@node-c11b-043.kathleen.uc   160
 146862 2.25862 REMD       xxxxxx      Rq    03/26/2021 06:24:39                                  160
 146911 2.19397 REMD       xxxxxx      Rq    03/26/2021 11:37:20                                  160
 146768 0.00000 REMD       xxxxxx      hqw   03/13/2021 14:47:35                                  160
 146769 0.00000 REMD       xxxxxx      hqw   03/13/2021 14:47:35                                  160
 146770 0.00000 REMD       xxxxxx      hqw   03/13/2021 14:47:36                                  160

The first element of each line is the job ID. Is there a way to show the lines for a particular range of jobs, e.g. how to only show the jobs from 146868 to 146927? It seems that

is needed.

I need to manage multiple cluster nodes from my PC. The catch is that the cluster nodes are accessible only from a remote PC. So first I have to ssh to the remote PC, and then ssh to a cluster node from the remote PC. If I had direct access to the nodes, I would use something like clusterssh to manage all the nodes simultaneously. Is there a tool similar to clusterssh that I can use in my situation?

Normally when I have my text editor open in terminal, I use the key combination CTRL+K+V to move to the end of the file CTRL+K+V does not work when I am in the "main cssh input field", (ie when i am typing in multiple terminals). But if I select one of the open terminals directly, then CTRL+K+V works. So I assume, when I use it in the main input field, it gets "eaten". Ideally, I would like cssh to pass all key combination to the terminals. I don't need cssh to use its own key shortcuts. Is there somewhere where I can configure / disable key shortcuts used by cssh, so that all are passed to terminals ? Another problem is, when I press F10, it is not passed to the terminals, but opens the menu of cssh. This is pure evil. The cssh window already has menu, where I can click with mouse and select items. Who needs to sacrifice F10, for such stupid thing? Is it possible to disable **all** keys in cssh, so that all are passed to the terminals ?

I have 3 monitors, one big and two small. When I start cssh, the xterm windows cover all 3 monitors. That would be fine, if the xterm windows did not overlap the monitor boundaries. But unfortunately, what I get is basically xterm window where one half is on monitor 1 and the other half on monitor 2. I would like to restrict cssh to only use one monitor. I am using openbox window manager, and I tried adding the following to my xml config: yes yes 2 Using the xml config, I can change details such as `, but 2` has no effect. Is there a way to tell cssh to only use monitor 2 ?

clusterssh (cssh) can pass options to ssh: cssh -o "-o ConnectTimeout=1" myserver.mydomain.org I am trying to pass ProxyCommand option to ssh, but it does not work. With plain ssh, I can use this: ssh -o ProxyCommand='ssh root@1.2.3.4 nc %h %p' behind-fw.mydomain.org which will connect to server non-routable.mydomain.org via 1.2.3.4. But when I try to use this with cssh: cssh -o "-o ProxyCommand='ssh root@1.2.3.4 nc %h %p'" non-routable.mydomain.org I get this error: Cannot open pipe for reading when talking to non-routable.mydomain.org : Interrupted system call How can I pass ProxyCommand to cssh ?

I'm trying to submit a job on a cluster via *qsub*, but it gets stuck in state "*Eqw*" with error message: $ qstat -j 466 | grep error error reason 1: 09/18/2015 17:12:32 [1125:3453]: error: can't chdir to /export/home/rafaelmf: No such file or direct I'm just using a test.sh script with echo "Hello World" output so I can try adding or removing options like *-cwd*, *-j*, *-o*. But nothing works. *qstat* also shows: sge_o_home: /export/home/rafaelmf sge_o_workdir: /state/partition1/home/rafaelmf and I know *export* is a link to *state/partition1/*. Also, I don't have root access, and all is done with *ssh*. So, does anyone know how to deal with such error (without *sudo*)?

I am running a large simulation on a computer cluster using 50 compute nodes. This solver uses a data structure which grows on the fly and (very) differently for each node. I need to make sure the memory used does not grow beyond each node's memory limit. So far, I am doing it in the most inefficient way: I have one terminal tab open for each node and run top to check the % memory used. Is there a way I can do it with a script? The idea would be to ssh on each node and store the memory usage, ssh to the next, etc...

I use the following code to submit a job on a cluster, but I don't know what these code means. Can some one explain me what the following code means if possible line-by-line. #!/bin/bash #PBS -N NAME_OF_JOB #PBS -l nodes=1:ppn=20 #PBS -l matlab_user=1 #PBS -l matlab_lic=20 #PBS -l min_walltime=1:00 #PBS -q small #PBS -S /bin/bash ##PBS -V ##PBS -m abe #PBS -j oe # cd $PBS_O_WORKDIR cat $PBS_NODEFILE export PATH=/opt/software/matlabr2014a/mdcs/bin:$PATH matlab -nodisplay -r "code1" -logfile code1.log Thanks

(directory=$(python -c 'import read_params; print(read_params.get_directory())') nsrc=$(cat $directory/master.pixels|wc -l) tail $directory/forward_src0{1..$nsrc}_ls00/out_data_forward) What the above does is that a simulation is running (what that is is not important) and the above command tracks it, by printing out the files in the directory: - ..../forward_src01_ls01/out_data_forward - ..../forward_src01_ls02/out_data_forward - ..../forward_src01_ls03/out_data_forward - ..../forward_src01_ls04/out_data_forward and so on till src_08. It runs in some linux systems without error. But in my system, it throws the below error: [prasad@daahpc1 sparc_flows]$ (directory=$(python -c 'import read_params; print(read_params.get_directory())') > nsrc=$(cat $directory/master.pixels|wc -l) > tail $directory/forward_src0{1..$nsrc}_ls00/out_data_forward) tail: cannot open /scratch/prasad/sparc_test/forward_src0{1..8}_ls00/out_data_forward' for reading: No such file or directory To be clear, these files do exist... I can open them individually, but not in a batch.

In my cssh config file, I have: terminal_font=6x13 this default font is too small for me. man cssh says, I can specify other font size with the -f argument, while using "standard X font notation". [Here](https://en.wikipedia.org/wiki/Fixed_(typeface)) is a wikipedia page containing some other fonts types. Unfortunately, none of these work for me: $ cssh host1 host2 -f "7x14" Fatal: Unrecognised font used (7x14). Please amend $HOME/.clusterssh/config with a valid font (see man page). Why does "7x14" not work? How can I use bigger fonts with cssh ?

I have a large script which relies on input arguments (with getopts). One of these arguments is a directory containing files (all named *bam) This script has 2 parts: - Part1: based on input *bam files, calculate one specific number. To be clear, the result is one single number, NOT one number per file. - Part 2: using the number found in part1, perform a series of operations on each *bam file. Now, originally, part1 was very quick, computationally speaking. So my setup was: - Run script on terminal: bash script.sh - Within the script.sh, for part2, make a HPC job submission for each file However, now that I need to analyze many more files than originally planned, I am realising that Part1 will also be computationally heavy - I therefore need to also run this on the HPC. So my question is: - Is it possible to submit an HPC job which submits jobs in it? - In other words, can I submit script.sh as a job and and still have it submit jobs in its part2? To be clear, here is an example of what my script might look like: #!/usr/bin/bash # PART 0: accept all input arguments USAGE() { echo "Usage: bash $0 [-b ] [-o ] [-c ]" 1>&2; exit 1; } if (($# == 0)); then USAGE fi # Use getopts to accept each argument while getopts ":b:o:c:h" opt do case $opt in b ) BAMFILES=$OPTARG ;; o ) OUTDIR=$OPTARG ;; c ) CHROMLEN=$OPTARG ;; h ) USAGE ;; \? ) echo "Invalid option: -$OPTARG exiting" >&2 exit ;; : ) echo "Option -$OPTARG requires an argument" >&2 exit ;; esac done # PART1: calculate this unique number NUMBER=0 for i in $(ls $BAMFILES/*.bam) do make some calculations on each file to obtain a number ... keep only the smallest found number and assign its value to $NUMBER done echo "Final number is ${NUMBER} " # PART2: Using $NUMBER that we found above, submit a job for each *bam file for i in $(ls $BAMFILES/*bam) do if [ ! -f ${OUTDIR}/${SAMPLE}.bw ]; then command=" command -options -b $NUMBER $i" echo $command | qsub -V -cwd -o $OUTDIR -e $OUTDIR -l tmem=6G -l h_vmem=6G -l h_rt=3600 -N result_${SAMPLE} fi done

I have read the question [cluster-ssh: specify terminal font](https://unix.stackexchange.com/q/230106/10489) and confirmed that my configuration works fine with valid non-Freetype font names as passed to, e.g., xfontsel. However, I would like to use Freetype fonts instead. With urxvt I specify my font as xft:Source Code Pro:size=9 in my X resources; the same specification works fine when given directly with the -fn command line option. However, if I use the same specification with the -f option to cssh or the terminal_font setting in .clusterssh/config I get an error message: Fatal: Unrecognised font used (xft:Source Code Pro:size=9). Please amend $HOME/.clusterssh/config with a valid font (see man page). How can I specify this font to ClusterSSH? (For those whom might be helped by a look at the source code, the error originates in the get_font_size() function of /usr/share/perl5/App/ClusterSSH.pm on my system; I suspect it's the $xdisplay->QueryFont($font); call that's failing.)

Is there any possibility to use all completed hosts with cssh? E.g: cssh www* This should establish three connections to www1, www2, www3 If I press Tab after www I get all three possibilities.

I want to add a large cluster with many consecutive IP addresses to my ~/.clusterssh/clusters. Is there an easy way to do this?

I have a cluster configured in my .csshrc config file, with several nodes. I want that when I open this cluster, a given command is run on each node automatically. The command is different for each node. Actually, the commands are very similar for each node, only a few characters change from node to node. Is there a way to set this up?

I would like cssh to use terminator terminal emulator, instead of the default xterm. In ~/.clusterssh/config I have changed the following line: #terminal=/usr/bin/xterm terminal=/usr/bin/terminator But when I start cssh I get following error: Opening to: server1 server2 Usage: terminator [options] terminator: error: no such option: -g Usage: terminator [options] terminator: error: no such option: -g Cannot open pipe for reading when talking to server1 : Interrupted system call Apparently, cssh is passing some parameters to xterm, which terminator does not recognize. How can I fix this problem ?

I am trying to install two node cluster for web server. We have super mico hardware HA BOX. It has 5 disk. one disk is used for node1 os installation and another disk is for node2 OS. Remaining disk is for quorum and shared storage. My problem is in node 1 it is detected as a /dev/sda and in node 2 it is detected has /dev/sdf but uuid is same for both the devices. Node1 lsblk output sdg 8:96 0 3.8G 0 disk sdb 8:16 0 34.2G 0 disk ├─sdb1 8:17 0 512M 0 part ├─sdb2 8:18 0 29.3G 0 part └─sdb3 8:19 0 4.4G 0 part sdd 8:48 0 34.2G 0 disk ├─sdd1 8:49 0 512M 0 part /boot └─sdd2 8:50 0 29.3G 0 part / sde 8:64 0 34.2G 0 disk └─mpathc (dm-3) 253:3 0 34.2G 0 mpath sdf 8:80 0 68.4G 0 disk └─sdf1 8:81 0 68.4G 0 part └─nfs-lvnfs (dm-1) 253:1 0 50G 0 lvm sda 8:0 0 137G 0 disk └─mpathe (dm-0) 253:0 0 137G 0 mpath └─mpathep1 (dm-7) 253:7 0 137G 0 part node2 lsblk output NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT sdg 8:96 0 3.8G 0 disk sda 8:0 0 34.2G 0 disk ├─sda1 8:1 0 512M 0 part /boot ├─sda2 8:2 0 29.3G 0 part / └─sda3 8:3 0 4.4G 0 part /swap sdc 8:32 0 34.2G 0 disk ├─sdc1 8:33 0 512M 0 part └─sdc2 8:34 0 29.3G 0 part sdd 8:48 0 34.2G 0 disk └─mpathb (dm-1) 253:1 0 34.2G 0 mpath sde 8:64 0 68.4G 0 disk └─sde1 8:65 0 68.4G 0 part sdf 8:80 0 137G 0 disk └─mpathd (dm-3) 253:3 0 137G 0 mpath └─mpathdp1 (dm-8) 253:8 0 137G 0 part sr0 11:0 1 1024M 0 rom blkid in node1 /dev/sdd2: UUID="69480877-c022-48af-8534-a3f91a587805" TYPE="ext4" /dev/sdb1: UUID="a59262f8-b37b-4c50-ba13-bbfb928f4fca" TYPE="ext4" /dev/sdb2: UUID="c6741a56-5b4d-4061-a20b-71f4c4d348ac" TYPE="ext4" /dev/sdb3: UUID="ee654c6e-232c-4b1b-bcda-233502dcce31" TYPE="ext4" /dev/sdd1: UUID="06f5a70b-135a-4e3b-8945-f7747fd5c133" TYPE="ext4" /dev/sdf1: UUID="GtrjBM-P2U9-VFcK-yvwd-JTmZ-1esg-dM3MZT" TYPE="LVM2_member" /dev/mapper/mpathbp1: UUID="a59262f8-b37b-4c50-ba13-bbfb928f4fca" TYPE="ext4" /dev/mapper/mpathbp2: UUID="c6741a56-5b4d-4061-a20b-71f4c4d348ac" TYPE="ext4" /dev/mapper/mpathbp3: UUID="ee654c6e-232c-4b1b-bcda-233502dcce31" TYPE="ext4" /dev/mapper/mpathep1: UUID="f6a58cf7-b39b-489c-add5-444ceab8af9d" TYPE="ext4" blkid in node2 [root@simha2 ~]# blkid /dev/sda2: UUID="c6741a56-5b4d-4061-a20b-71f4c4d348ac" TYPE="ext4" /dev/sda1: UUID="a59262f8-b37b-4c50-ba13-bbfb928f4fca" TYPE="ext4" /dev/sda3: UUID="ee654c6e-232c-4b1b-bcda-233502dcce31" TYPE="ext4" /dev/sdc1: UUID="06f5a70b-135a-4e3b-8945-f7747fd5c133" TYPE="ext4" /dev/sdc2: UUID="69480877-c022-48af-8534-a3f91a587805" TYPE="ext4" /dev/sde1: UUID="GtrjBM-P2U9-VFcK-yvwd-JTmZ-1esg-dM3MZT" TYPE="LVM2_member" /dev/mapper/mpathcp1: UUID="GtrjBM-P2U9-VFcK-yvwd-JTmZ-1esg-dM3MZT" TYPE="LVM2_member" /dev/mapper/mpathap1: UUID="06f5a70b-135a-4e3b-8945-f7747fd5c133" TYPE="ext4" /dev/mapper/mpathap2: UUID="69480877-c022-48af-8534-a3f91a587805" TYPE="ext4" /dev/mapper/mpathdp1: UUID="f6a58cf7-b39b-489c-add5-444ceab8af9d" TYPE="ext4" I want use uuid of UUID="f6a58cf7-b39b-489c-add5-444ceab8af9d" in cluster. In luci web console under resource tab i have given above uuid to autodetect and mount files system in both the nodes since we can use Device, FS Label, or UUID. I given UUID . I am starting service group in luci console i am getting error could't start clusvcadm . Below is rgmanger.log May 31 16:53:29 rgmanager Reconfiguring May 31 16:53:29 rgmanager Loading Service Data May 31 16:53:31 rgmanager Stopping changed resources. May 31 16:53:31 rgmanager Restarting changed resources. May 31 16:53:31 rgmanager Starting changed resources. May 31 16:54:10 rgmanager Stopping service service:Web directory May 31 16:54:10 rgmanager [fs] stop: Could not match "f6a58cf7-b39b-489c- add5-444ceab8af9d" with a real device May 31 16:54:10 rgmanager Service service:Web directory is disabled May 31 16:54:11 rgmanager [fs] start_filesystem: Could not match "f6a58cf7- b39b-489c-add5-444ceab8af9d" with a real device May 31 16:54:14 rgmanager [fs] start_filesystem: Could not match "f6a58cf7- b39b-489c-add5-444ceab8af9d" with a real device May 31 16:54:17 rgmanager [fs] start_filesystem: Could not match "f6a58cf7- b39b-489c-add5-444ceab8af9d" with a real device May 31 16:54:20 rgmanager start on fs "Web document root" returned 2 (invalid argument(s)) May 31 16:54:20 rgmanager #68: Failed to start service:Web directory; return value: 1 May 31 16:54:20 rgmanager Stopping service service:Web directory May 31 16:54:20 rgmanager [fs] stop: Could not match "f6a58cf7-b39b-489c- add5-444ceab8af9d" with a real device May 31 16:54:20 rgmanager stop on fs "Web document root" returned 5 (program not installed) May 31 16:54:20 rgmanager #12: RG service:Web directory failed to stop; intervention required May 31 16:54:20 rgmanager Service service:Web directory is failed May 31 16:54:20 rgmanager #2: Service service:Web directory returned failure code. Last Owner: simha1 May 31 16:54:20 rgmanager #4: Administrator intervention required. Attached cluster screen shot for reference. Even i tried with assigning label name to filesystem not working and also i tried with filesystem mount like /dev/mapper/mpathep1 then it is working for node1 and when i trun off node1 is not starting since /dev/mapper/mpathep1 is not present in node Os Redhat 6.5. Please help me