Installation notes - 1.Oct.2021¶
The infrastructure has been fully renewed adapting to new situation e.g. Malt. Some of the applications notably Ms Visio (.vsd or .vsdx) or Ms Project (.mpp) are no longer available at the converter servers. This has been discussed with Malt office responsibles and it was the final outcome.
For convenience, most of the software done in this exercise has been copied at \\cern.ch\dfs\Services\conversion\Software\Neevia\WorkerNodeInstallation-Oct2021
.
- Install Chrome, Sublime Text, Moba Xterm, Git, Python (Custom installation: all users & add to environment variables, install as admin and add:
C:\Program Files\Python39\
andC:\Program Files\Python39\Scripts\
andC:\Users\cdsconv\AppData\Roaming\Python\Python39\Scripts
as environment variables) - In development machine, also install VStudio Professional
Setting up Python:
--virtualenv as admin
C:\>python.exe -m pip install --upgrade pip
pip install virtualenv
C:\doconverter\doconverter>.\venv\Scripts\activate.bat
pip install pywin32
pip install -r requirements.txt
--as user
C:\doconverter\doconverter>virtualenv venv
--modify activate.bat and add
set "PYTHONPATH=C:\doconverter\doconverter"
Mount Y: drive:
- production: \cernbox-smb.cern.ch\eos
- qa: \cernbox-smb.cern.ch\eos\user\c\cdsconv
and create directories for the server: /var and /archive.
Copy all directories but the software one from a production or qa machine to c:\doconverter.
Adapt doconverter.ini
as needed.
Install telegraf¶
Get the latest at https://portal.influxdata.com/downloads/. Copy the telegraf.conf from a different server. Then
# run a single telegraf collection, outputting metrics to stdout
telegraf --config telegraf.conf --test
# install telegraf service
telegraf --service install --config "C:\Program Files\Telegraf\telegraf.conf"
Go to services.msc and start it.
Instal filebeat, download it from Elasticsearch site, copy filebeat.yml from prod or qa to the appropiate machines.¶
(venv) C:\Program Files\Elastic\Beats\7.14.0\filebeat>filebeat test config -c C:\programdata\Elastic\Beats\filebeat\filebeat.yml
Config OK
(venv) C:\Program Files\Elastic\Beats\7.14.0\filebeat>filebeat test output -c C:\programdata\Elastic\Beats\filebeat\filebeat.yml
logstash: logstash-icappsqa.cern.ch:5044...
connection...
parse host... OK
dns lookup... OK
addresses: 188.184.86.119
dial up... OK
TLS... WARN secure connection disabled
talk to server... OK
# Create the new service.
New-Service -name filebeat `
-displayName Filebeat `
-binaryPathName "`"C:\Program Files\Elastic\Beats\7.14.0\filebeat\filebeat.exe`" -environment=windows_service -c `"C:\programdata\Elastic\Beats\filebeat\filebeat.yml`" -path.home `"C:\programdata\Elastic\Beats\filebeat`" -path.data `"C:\programdata\Elastic\Beats\filebeat\data`" -path.logs `"C:\programdata\Elastic\Beats\filebeat\logs`" -E logging.files.redirect_stderr=true"
Install tesseract¶
https://medium.com/@ahmetxgenc/how-to-use-tesseract-on-windows-fe9d2a9ba5c6, find binaries for windows at https://github.com/UB-Mannheim/tesseract/wiki, Install the exe file located at \cern.ch\dfs\Services\conversion\Software\OCR\tesseract. For Version 5, Chose to install for everyone and add FRench as additinal language.
At Control Panel -> System -> Edit Environment vairables: system and add at Path: "C:\Program Files\Tesseract-OCR"
Autocad¶
Based on stats from last 2 years it has not been used, considering: https://cern.service-now.com/service-portal?id=kb_article&n=KB0006310 the cost is excessive for an app that it's not used.
A free option proposed by Neevia support was to install Design Review and TrueView 2018, both packages can be found at: https://www.autodesk.com/products/design-review/download and here: https://download.autodesk.com/esd/dwgtrueview/2018/DWGTrueView_2018_ENU_64bit.sfx.exe.
Create shedule tasks from a prod/qa server.¶
Copy folder C:\CERNTools
from an existing server. Export tasks from Task Scheduler in origin server and re-import them in the target server.
doconverter software¶
c:\
mkdir doconverter
cd doconverter
git clone https://github.com/CERNCDAIC/doconverter.git
- Copy all other directories in a production machine e.g.: etc, config, files,...
- Create archival directore at c:\users\cernbox\cernboxprod, actual year should be just enough. This is used by the schedule tasks.
- Create directories in the EOS folders
- Configure the frontend on openshift (dev and prod) to point to the right servers.