I expect you to have no trouble, but I think this will be the link to download for that version that was oh so briefly run on the main site:

https://www.cpdn.org/cpdnboinc/applications/openifs_2.19_x86_64-pc-linux-gnu

Great! Thank you very much. Unless they add some spooky extra libraries, this will run on my machine. So I expect it will run on all RHEL 8 and CentOS 8 systems as well.
If it becomes a real memory hog, I could increase my RAM to 96 GBytes or 128 GBytes, but the processor cache will have to remain at 16,896 KBytes.

$ file openifs_2.19_x86_64-pc-linux-gnu 
openifs_2.19_x86_64-pc-linux-gnu: ELF 64-bit LSB executable, x86-64, version 1 (SYSV), dynamically linked,
     interpreter /lib64/ld-linux-x86-64.so.2, for GNU/Linux 2.6.32,
     BuildID[sha1]=9daea76ce24beb521ecb0187fab00ffda9e11718,
     with debug_info, not stripped

Computer 1511241
CPU type 	GenuineIntel Intel(R) Xeon(R) W-2245 CPU @ 3.90GHz [Family 6 Model 85 Stepping 7]
Number of processors 	16
Operating System 	Red Hat Enterprise Linux 8.5 (Ootpa) [4.18.0-348.12.2.el8_5.x86_64|libc 2.28 (GNU libc)]
BOINC version 	7.16.11
Memory 	62.4 GB
Cache 	16896 KB

$ ldd openifs_2.19_x86_64-pc-linux-gnu 
	linux-vdso.so.1 (0x00007fff8c3b0000)
	libstdc++.so.6 => /lib64/libstdc++.so.6 (0x00007f2031d6b000)
	libm.so.6 => /lib64/libm.so.6 (0x00007f20319e9000)
	libgcc_s.so.1 => /lib64/libgcc_s.so.1 (0x00007f20317d1000)
	libpthread.so.0 => /lib64/libpthread.so.0 (0x00007f20315b1000)
	libc.so.6 => /lib64/libc.so.6 (0x00007f20311ec000)
	/lib64/ld-linux-x86-64.so.2 (0x00007f2032100000)

---

I am getting quite a few fails of late, with the error "CPDN process is not running", being a bit of a Linux n00b can someone let me know how to check if it is running / how to restart.
It was working fine for the last lot of units I got, not sure if OS updates have borked things?
Running Linux Mint in a VM.

https://www.cpdn.org/results.php?userid=2003084&offset=0&show_names=0&state=6&appid=

It might be not enough memory.
Climate models are BIG. Allow perhaps 2 gigs of ram per model.

I am getting quite a few fails of late, with the error "CPDN process is not running", being a bit of a Linux n00b can someone let me know how to check if it is running / how to restart.
It was working fine for the last lot of units I got, not sure if OS updates have borked things?
Running Linux Mint in a VM.

The high resolution hadam4h '216' models here use 1.5GB per model versus 684MB for the hadam4 '144' models. Here, three 216's and one 144 are currently filling up the 8GB ubuntu VM memory to leave just 140MB 'Free'.

Looking at your Windows host, 32 Processor with 32GB of RAM; and the Mint VM with 24 processors in 16GB of virtual RAM ... I'd suggest that you limit the CPU count for the Mint VM in VirtualBox (system settings) to 6 CPUs for 16GB of VM memory; or perhaps 8 CPUs for 24GB of VM memory. You'll need to power down the VM to make the VBox setting changes. You can further control the CPU granularity in BOINC manager: Options, Computing preferences.

It's also important to turn off (pause) Windoze updates for as long as Doze lets you, as updates will crash models when you least expect it.

---

Its a Linux host.

Any way to limit the number of jobs it does at once? I run other projects on the VM as well and don't want to lower the thread count.

Its a Linux host.

Any way to limit the number of jobs it does at once? I run other projects on the VM as well and don't want to lower the thread count.

I think there may be an option in cc_config.xml that lets you limit the number of tasks for an individual project. I couldn't be sure however. If you do decide to edit this file, do make a backup of it first. Another possibility is client_state.xml. I have edited that one in the past without problems.

Edit: I think this has been asked on the BOINC message boards in the past. You don't have to sign up to search them.

The full list of options is at https://boinc.berkeley.edu/wiki/Client_configuration#Options.

You might be thinking of

<ncpus>N</ncpus>
Act as if there were N CPUs; e.g. to simulate 2 CPUs on a machine that has only 1. Zero means use the actual number of CPUs. Don't use this to limit CPU usage; use computing preferences instead.

Doesn't limit the number of tasks directly, but can limit the number of devices available to run them - which amounts to the same thing.

Its a Linux host.

Any way to limit the number of jobs it does at once? I run other projects on the VM as well and don't want to lower the thread count.

max_concurrent in app_config.xml but be aware that this can lead to runaway downloads if you’re unlucky.

this can lead to runaway downloads

Should be fixed in BOINC client 7.20.

See:
https://github.com/BOINC/boinc/issues/4322

The runaway work fetch seems to be only caused by max_concurrent not project_max_concurrent. Thus a very simple app_config as in the following example will run 4 CPDN tasks (mix of N144, N216, whatever is available) at a time and avoid potential runaway work fetch. Place it in the CPDN project folder and run the "Read config files" menu option/command.

<app_config>
   <project_max_concurrent>4</project_max_concurrent>
</app_config>

Ok, looks like not possible to put libstd++ in with task on met office models. Shucks!

Previous text deleted.

A new version of UBUNTU (22.04 Jammy Jellyfish) is apparently available. Any ideas as to which libraries will be required (before i upgrade and find its not working!)?

A new version of UBUNTU (22.04 Jammy Jellyfish) is apparently available. Any ideas as to which libraries will be required (before i upgrade and find its not working!)?

I think the same as for 20.10 still works. I didn't do a clean install but just upgraded so can't be 100% sure.

    sudo  apt-get  install  lib32ncurses6  lib32z1  lib32stdc++-8-dev

It worked for 21.10

Thanks. I'll let a couple more tasks finish then do the upgrade. At most I'll lose a couple of tasks.

Thanks. I'll let a couple more tasks finish then do the upgrade. At most I'll lose a couple of tasks.

If running the upgrade rather than doing a clean install (I do the latter every couple of years which clears out a lot of the rubbish.) All the needed libraries should already be there.

I'll just try it an see if it works then.

It worked for 21.10

Hmm. Yesterday, I tried to make a new Jammy Jellyfish ubuntu 22.04 VM on our latest i7-8500 Optiplex PC, WIN10 with VirtualBox. The install wouldn't behave when I tried to install the 32 bit libraries. I've faffed around with SUDO commands and library versions without success. Any good ideas are welcome.?
Today, I reverted to exporting the ubuntu 20.04 VM appliance, copying the export file to the new PC, and importing it onto the new VB setup. That worked easily. I've done the same export/copy/import for Apple Mojave .

---

BTW, the imported Mojave appliance is crashing :((

---

Upgrade worked fine but couldn't find libs32stdc++-8-dev to install. Last 2 tasks completed OK. Now shutdown to conserve power (big bills) and waiting for new work.