Do Linux WU's run faster than Windows?

In this project, The types of program that run on each OS are completely different.
You can't compare speed.

It's like asking about cars V trains, or elephants V chickens.

There are a few other people around who will remember the days when all tasks were multiplatform who may have better memories than I do about the difference between operating systems then. (It was in my early days of using BOINC and I knew a lot less about all of the different variables involved than I do now!
--------------------------------------------
Early days, like 8086 and 8088 of Texas Instruments. Command Line GW-Basic, then along came IBM PC in 1980 if my memory serves me correctly. Problem is, our memories are all shot up plus some flavour of windows came around in the late '80s and spoilt me silly. I hope it all comes back.
Thank you.

There are a few other people around who will remember the days when all tasks were multiplatform who may have better memories than I do about the difference between operating systems then. (It was in my early days of using BOINC and I knew a lot less about all of the different variables involved than I do now!
--------------------------------------------
Early days, like 8086 and 8088 of Texas Instruments. Command Line GW-Basic, then along came IBM PC in 1980 if my memory serves me correctly. Problem is, our memories are all shot up plus some flavour of windows came around in the late '80s and spoilt me silly. I hope it all comes back.
Thank you.

I think I had graduated to a 286 or possibly even a 386sx by the time I first ran CPDN work. Checking would be difficult as I lost the email I had then a long while ago and the account it was linked to here will have been deleted under data protection policies by now.

---

Please do not private message myself or other moderators for help. This limits the number of people who are able to help and deprives others who may benefit from the answer.

This question really is about different VMs. I started using VMware (Ubuntu 18.04 LTS) but have recently switched to Vbox (Ubuntu 20.04) and all tasks are running more slowly. Anyone got any ideas? (Sorry if this is off topic).

I don't know the fine details of these things, but if it was plain vanilla Linux, I'd say more overheads.
Any programs that allow you to see what is running, or perhaps more importantly, how fast processes are running?

Ah! What time period are you talking about; microseconds, minutes, hours?
Could also be write/read to/from HD.

This question really is about different VMs. I started using VMware (Ubuntu 18.04 LTS) but have recently switched to Vbox (Ubuntu 20.04) and all tasks are running more slowly. Anyone got any ideas? (Sorry if this is off topic).

I had the same issue as you did. Used VMWare and decided to try VBOX. Noticeably slower even with the same distribution and even in the hardinfo benchmark. I'm not sure why the difference, but my VBOX experiment didn't last long.

I have only used VBox from Linux at a time when CPDN didn't have any work. If no one here has the right experience to answer the question as to why one way of running tasks in a virtual machine is so much slower than another it may be worth asking over in the BOINC forums. If doing so I would first want to confirm that it is true with other projects and not just with CPDN.

Disclaimer, I have never tried running CPDN in a VBox client, so what follows may be irrelevant, but it does relate to VBox performance issues.

I used to run 32-bit WCG tasks in a VBox client for throughput experiments (and usually the 32-bit tasks in the VM finished slightly quicker than similar 64-bit tasks on the host) I still use VBox for other, non BOINC-related, tasks such as a Windows VM for my camera update software(!) and a Linux VM for home banking and such like.

A couple of observations about VirtualBox: it seems that whenever I did a significant version upgrade (rather than a maintenance update) something would need re-configuring. If lucky, it would just be the video or sound; however, it was often something that affected performance!

Firstly, if you don't install up-to-date Guest Additions, performance or behaviour may suffer.

Then there's the VBox version... One major update I did turned off the user of KVM paravirtualization, for instance. Performance hit! And somewhere around my move from XUbuntu 18.04 to 20.04 (host and client, in this particular case), it turned out that one of my VMs was not using host cache for disk I/O - that VM was running like a slug until I found that issue via the VBox online forums.

My apologies if none of the above is news, but as I don't see massive performance hits if I time some of my home-grown non-interactive programs in a VM and on the host of that VM (trying to ensure nothing else is busy at the time), I wonder if some of it is "tuning" related.

I have never tried using Windows as a host, by the way; the above is all Linux-hosted. Someone hoating on Windows might know whether something akin to the KVM or host-cache might be an issue there...

Cheers - Al.

[Edited to add the rider about Windows hosts...]

---

"What time period are you talking about; microseconds, minutes, hours?"

Tasks from batch 887 were running at about 35sec/ts on VMware whereas tasks from batch 889 are running at about 48sec/ts on VBox, so quite a big difference. I am trying VBox because VMware sometimes freezes requiring a "hard" termination and reboot of the computer.

I was advised to post in this forum regarding my problem. I hope it's OK to post to this thread regarding this same subject; if not, I apologize, and I'll start a new thread.

I'm unable to receive any WUs on my Linux box (Mint 18.3). The event log says "Scheduler request completed: got 0 new tasks" but gives no clue as to why. A while ago I updated to 64 bit, but in my ignorance didn't think to check compatibility and didn't install the 32-bit libraries. I resolved that this morning, but am I now on a bad-boy list ?

Additional info: I just swapped the mobo/proc on this machine to a Reizen 5, 16G ram. Perhaps I should disconnect and reattach the project?

Thanks.

There's an ongoing problem with this.
It's just some computers.

The only message is something like " No Tasks sent", no matter which debug flags are set.

And I'm starting to suspect that it's a server bug, which only affects AMD computers, perhaps only Ryzen.

I think that it's time to get the project people involved.

Les, to address both your latest post, and to provide an update to my post just preceding it...

I have just (as in 5 minutes ago) removed and reconnected to the cpdn project using the newly-upgraded Ryzen machine I spoke of earlier. BOINC immediately downloaded and began running a WU. So far I'm up and running at... just rolling over 8 minutes. Projected completion for this WU is in 24 days 0 hrs...
Application is "UK Met Office HadAM4", if that is of any value...

Les, to address both your latest post, and to provide an update to my post just preceding it...

I have just (as in 5 minutes ago) removed and reconnected to the cpdn project using the newly-upgraded Ryzen machine I spoke of earlier. BOINC immediately downloaded and began running a WU. So far I'm up and running at... just rolling over 8 minutes. Projected completion for this WU is in 24 days 0 hrs...
Application is "UK Met Office HadAM4", if that is of any value...

Glad to see the problem resolved :-)

Well it is for me, thanks, but the issue of uncertainly still l remains: I see that folks have been confused about this for quite a while. I don't really know why the server stopped working when it did; there was no message that told me so, much less why, and there was no clue on my stats or config pages either: that's tantamount to passive aggressive behavior. I do look at the notices whenever they pop up, and at the Event Log whenever something appears to be amiss, but there was no clue in either place. I read the laments on the threads that people don't look to make sure that the Linux versions of the various apps are working properly, but the communication isn't the most effective.

Anyway, bottom line for me is that removing and reconnecting to the project did the trick.

Thanks for your help, Les, to me and to all of us willing-but-not-always-fully-competent worker bees :-)

I ran into some issues on latest linux Mint 20.2 update, which is based on Ubuntu Focal Fossa (20.04 LTS). I managed to track it down to some problem with the versioning of libc6, which is reported here: https://bugs.launchpad.net/ubuntu/+source/glibc/+bug/1926918
Eventually, I managed to get it working with;

sudo  apt-get -V  install  lib32ncurses6  lib32z1  lib32stdc++-7-dev lib32tinfo6 libc6-i386=2.31-0ubuntu9.2 lib32gcc-7-dev libc6=2.31-0ubuntu9.2 libstdc++-7-dev libc6-dev=2.31-0ubuntu9.2 libc-dev-bin=2.31-0ubuntu9.2

Hope it helps somebody :-)

@WynX

Are you saying that something broke going from Mint 20.1 to 20.2, or that you had to install those other libraries with a new install of Mint 20.2? Just wondering since the PC I'm running this summer is running Mint 20.1 and when I freshly installed that version, the command line below for Ubuntu 20.04 LTS worked fine.

sudo  apt-get  install  lib32ncurses6  lib32z1  lib32stdc++-7-dev

Edit...also, how could you tell there was a problem if it was a simple update from 20.1 to 20.2? Did it give an error?

Thanks.

There are obsolete pinned threads and I'm wondering if this is one of them. How do I know if I actually need to do this? My WUs are running and not failing on Linux Mint 20.2 Ubuntu 20.04. I tried running the recommended command but the Synaptic Package Manager has now designated them Broken and cannot repair them.

aurum@Rig-34:~$ sudo  apt-get  install  lib32ncurses6  lib32z1  lib32stdc++-7-dev
Reading package lists... Done
Building dependency tree       
Reading state information... Done
Some packages could not be installed. This may mean that you have requested an impossible situation or if you are using the unstable distribution that some required packages have not yet been created or been moved out of Incoming.
The following information may help to resolve the situation:
The following packages have unmet dependencies:
 lib32ncurses6 : Depends: lib32tinfo6 (= 6.2-0ubuntu2) but it is not going to be installed
                 Depends: libc6-i386 (>= 2.7) but it is not going to be installed
 lib32stdc++-7-dev : Depends: lib32gcc-7-dev (= 7.5.0-6ubuntu2) but it is not going to be installed
                     Depends: lib32stdc++6 (>= 7.5.0-6ubuntu2) but it is not going to be installed
                     Depends: libstdc++-7-dev (= 7.5.0-6ubuntu2) but it is not going to be installed
 lib32z1 : Depends: libc6-i386 (>= 2.4) but it is not going to be installed
E: Unable to correct problems, you have held broken packages.

I ran into some issues on latest linux Mint 20.2 update, which is based on Ubuntu Focal Fossa (20.04 LTS). I managed to track it down to some problem with the versioning of libc6, which is reported here: https://bugs.launchpad.net/ubuntu/+source/glibc/+bug/1926918
Eventually, I managed to get it working with;

sudo  apt-get -V  install  lib32ncurses6  lib32z1  lib32stdc++-7-dev lib32tinfo6 libc6-i386=2.31-0ubuntu9.2 lib32gcc-7-dev libc6=2.31-0ubuntu9.2 libstdc++-7-dev libc6-dev=2.31-0ubuntu9.2 libc-dev-bin=2.31-0ubuntu9.2

Hope it helps somebody :-)

I tried your command and it seemed to work for the first 2 Linux computers but the third gave me this:

The following packages have unmet dependencies:
 lib32gcc-7-dev : Depends: libx32gcc-s1 (>= 7.5.0-6ubuntu2) but it is not going to be installed
                  Depends: libx32gomp1 (>= 7.5.0-6ubuntu2) but it is not going to be installed
                  Depends: libx32itm1 (>= 7.5.0-6ubuntu2) but it is not going to be installed
                  Depends: libx32atomic1 (>= 7.5.0-6ubuntu2) but it is not going to be installed
                  Depends: libx32asan4 (>= 7.5.0-6ubuntu2) but it is not going to be installed
                  Depends: libx32ubsan0 (>= 7.5.0-6ubuntu2) but it is not going to be installed
                  Depends: libx32cilkrts5 (>= 7.5.0-6ubuntu2) but it is not going to be installed
                  Depends: libx32quadmath0 (>= 7.5.0-6ubuntu2) but it is not going to be installed
E: Unable to correct problems, you have held broken packages.

I don't know what I'm doing but here's what I did. First try running this:

sudo  apt-get  install  lib32ncurses6  lib32z1  lib32stdc++-7-dev

If it doesn't work then I run this:

sudo  apt-get -V  install  lib32ncurses6  lib32z1  lib32stdc++-7-dev lib32tinfo6 libc6-i386=2.31-0ubuntu9.2 lib32gcc-7-dev libc6=2.31-0ubuntu9.2 libstdc++-7-dev libc6-dev=2.31-0ubuntu9.2 libc-dev-bin=2.31-0ubuntu9.2 --allow-downgrades -y

BOINC projects are still working so hopefully this did some good.

**** IMPORTANT ****

As of October 2019, all cpdn apps are still 32bit. Most/all 64bit Linux distributions do not by default have 32bit compatibility libraries installed. In order to get cpdn models to run on 64bit Linux, you must have these libraries installed. Below is guidance for later versions of Ubuntu and Redhat/CentOS to have compatibility with cpdn.
*********************************************************

Ah this explains why I haven't been able to get tasks... seemingly forever. Guess I should check the forums more than once every 5 years. Hopefully I'll be back to crunching this project soon!