Bug found in ems_prep: file path changed in Pacquire.pm

Questions and other topics related to UEMS 18.
Post Reply
Wilfred Janssen
Posts: 4
Joined: Wed Nov 23, 2016 10:12 am

Bug found in ems_prep: file path changed in Pacquire.pm

Post by Wilfred Janssen » Sat Feb 03, 2018 9:10 pm

Hello,

I'm trying to run the new UEMS code for a while now. I'm working on a system (Ubuntu 16.04) with the OS running on a 120GB SSD and a 2TB conventional harddisk for all my files.

Since WRF requires a lot of memory (in my case over 100GB), I installed the UEMS system on my conventional harddisk (which is mounted to the system). The path to my harddisk is now '/mnt/HDD/'. Now the problem: Installing UEMS on the harddisk is not any problem, but running is. I triple checked all environment variables, but the benchmark case kept telling me that the initialisation data files are not present (they were, obviously). Since then, I decided to check out the ems_prep code, and from there I ended up looking in the Uprep folder.

The first file I checked, was the Pacquire.pm file. With some print-commands I found out that the file path somehow has changed to /mnt/H27/.... This means that the 'DD' in the file's path has changed to a day number (27, later 28) in case of the benchmark case.

This may be a 'bug' which needs to be fixed in upcoming versions. My Perl-knowledge is too poor to fix it myself (unfortunately). Maybe there are more issues like this in the UEMS code (I don't know). I quick fixed the problem just by renaming the harddisk from HDD to something else without the 'DD' in it.

I hope this information will inform the code developer(s).

Last but not least, excuse me for my English language ;-)


Kind regards,

Wilfred Janssen

alfe
Posts: 86
Joined: Thu Nov 25, 2010 8:13 pm

Re: Bug found in ems_prep: file path changed in Pacquire.pm

Post by alfe » Sun Feb 04, 2018 11:39 am

Open a terminal window and type sysinfo. What does it say ?

Wilfred Janssen
Posts: 4
Joined: Wed Nov 23, 2016 10:12 am

Re: Bug found in ems_prep: file path changed in Pacquire.pm

Post by Wilfred Janssen » Mon Feb 05, 2018 8:30 pm

Starting UEMS Program sysinfo (V18.1.3) on Meteocomputer at Mon Feb 5 20:26:42 2018 UTC


☺ Collecting information about the machine known as "Meteocomputer"

System Information for Meteocomputer

System Date : Mon Feb 5 20:26:44 2018
System Hostname : Meteocomputer
System Address : None available

System OS : Linux
Linux Distribution : Ubuntu 16.04.3 LTS
OS Kernel : 4.10.0-28-generic
Kernel Type : 0-28-generic

Network Interface Information for Meteocomputer

Network Interface : enp2s0
Interface Address : 192.168.8.108
Address Resolves to : Meteocomputer
Interface State : Up

Network Interface : lo
Interface Address : 127.0.0.1
Address Resolves to : localhost
Interface State : Up

Processor and Memory Information for Meteocomputer

CPU Name : Intel(R) Core(TM) i7-4790K CPU @ 4.00GHz
CPU Microarchitecture : Haswell
CPU Type : x86_64
CPU Speed : 4.000 MHz

UEMS Determined Processor Count
Sockets : 1
Cores per Socket : 4
Total Cores : 4

UEMS.profile Specified Processor Count
Sockets : 1
Cores per Socket : 4
Total Cores : 4

Hyper-Threading : On Note: Attempting to use virtual "Hyper-threaded" CPUs while running the UEMS may result in a degradation in performance.

System Memory : 32 Gbytes

UEMS User Information for wilfred on Meteocomputer

Home Directory : /home/wilfred
User ID : 1000 (wilfred)
Group ID : 1000 (wilfred)
Home Directory Mount : Local
User Shell : /bin/bash
Shell Login Files : .bash_profile

UEMS Installation Information for Meteocomputer

UEMS Home Directory : /mnt/HDD/uems
UEMS User ID : 1000 (wilfred)
UEMS Group ID : 1000 (wilfred)
UEMS Home Mount : Local
UEMS Binaries : x64
UEMS Release : 18.1.3
UEMS WRF Release : 3.9.1

UEMS Run Directory : /mnt/HDD/uems/runs
UEMS Run Dir User ID : 1000 (wilfred)
UEMS Run Dir Group ID : 1000 (wilfred)
UEMS Run Dir Mount : Local

Run Dir Total Space : 885.35 Gb
Run Dir Space Used : 286.84 Gb
Run Dir Avail Space : 553.51 Gb



Your awesome UEMS Run Information party is complete - Mon Feb 5 20:26:44 2018 UTC

As the UEMS Genie says: "Think Globally, Model Locally!"





Seems all normal to me. However, changing the HDD path from 'HDD' to 'hdd' solved the problem. My hypothesis now is that capital double d's are changed to day numbers in the prep-routine. Maybe the same will happen with monhts (double m's) etc.

Post Reply