Giter Site home page Giter Site logo

docs's People

Contributors

aaronmbr avatar abigailrose avatar apertome avatar arlake228 avatar daldoyle avatar danielneto avatar dojosout avatar jwzurawski avatar laeti-tia avatar matejv avatar mfeit-internet2 avatar okjed avatar rockymm avatar schevali avatar soichih avatar sowmyab avatar szymontrocha avatar timchown avatar vvidic avatar

Stargazers

 avatar  avatar  avatar  avatar

Watchers

 avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar

docs's Issues

Full HA example of MA

It would be nice to add an example of a fully redundant, load balanced measurement archive to the clustering page. Right now it does a survey of the various options, but a single example would be good. The example should include:

  • A topology with 2 load balancers, two httpd servers running esmond, 2 cassandra servers and 2 PostgreSQL servers. Alternatively you could just have two load balancers and two more servers running everything else.
  • Setup a cassandra cluster
  • Create a master slave setup with postgreSQL with automatic failover (possibly using the load balancers)
  • Setup two web servers with esmond, both pointing at the clusters above
  • Setup two load balancers with HAProxy and keepalived
  • Optionally install HAproxy on each web server and use it to detect cassandra failures better. pycassa relies on timeouts, HAProxy would remove from the pool entirely. Note we would run this client side, not on cassandra servers

Update mesh config documentation

From @arlake228 on March 14, 2015 12:44

Original issue 1025 created by arlake228 on 2014-12-22T15:10:28.000Z:

A user writes:

=-=-=-=-

I have some things that are not working correctly on my mesh (traceroute and pingER tests are not working, owamp tests do work tho). Since this is broken on every single one of my mesh hosts (all running toolkit 3.4.1), I'm assuming there is something wrong with my mesh config.

Question: Where can I find a valid example mesh config file?
Per the perfSonar Mesh Config Instructions found here:
https://code.google.com/p/perfsonar-ps/wiki/MeshConfigurationInstallation
There is an example mesh config located at /opt/perfsonar_ps/mesh_config/doc/example.conf and also at:
http://perfsonar-ps.googlecode.com/git/perfSONAR_PS-MeshConfig/doc/example.conf

However the read_url and write_url values in these example files look wrong to me (they look like pre-esmond URLs) so I'm assuming (perhaps incorrectly) that these examples are out of date.

Can someone clarify if the above example mesh config files are valid or not, and if they are not, is there updated documentation somewhere on the recommended way to create a mesh config file in the correct format?

=-=-=-=-

They were pointed to this location, the wiki and RPM should be updated (or removed) so that we are consistent:

http://docs.perfsonar.net/multi_server_config.html#generating-the-configuration

The user goes on to say that in this example:

http://docs.perfsonar.net/multi_server_config.html#generating-the-configuration

does not have any "measurement_archive" sections included (where the read_url and write_url configs normally live). Is the "measurement_archive" config optional in the latest version, or is this example config incomplete?

Copied from original issue: perfsonar/project#1023

add pscheduler pages

Fill in the following files:

  • using_pscheduler.rst
  • config_pscheduler.rst
  • config_pscheduler_limits.rst

Remove mention of Ubuntu 12.04

Ubuntu 12.04 LTS reaches End of Life on April 28th 2017. It is probably better not to release perfSONAR packages for this distro nor support it with our 4.0 release.

If we decide this, we should remove all mention of Ubuntu 12.04 LTS from the documentation and the website.

Better document numeric types in meshconfig

From Soichi:

Lastly, the documentation doesn't explain if numeric fields can be integer or decimal (and if decimal, number of acceptable precision). For example, it looks like bucket_width/packet_interval can be decimal, but sample_count is an integer? If you could add more specific numeric specification, I can add appropriate validations on the UI to prevent user from entering invalid numbers.

make sure that manage_logs.rst is current

Mark:
I added pscheduler.log to this file, but I'm not sure if the description is accurate, or what other pscheduler logs should be worth mentioned. Tho maybe the logs should be made to match the description, and that info is typically quite useful.

Also, others should confirm that the components they are responsible for are accurate too.

Assigning this issue to everyone I think probably should take a look. Please add a comment when you confirm that your stuff is accurate.

Add troubleshooting guide

This a placeholder issue to collect any items for addition to a Troubleshooting document for all things installation, meshconfig, and MaDDash

Example:

  • check /etc/perfsonar/meshconfig-agent-tasks.conf for erroneous MA definitions and remove to cut down on log trash and processor use.

Document Simplestreamer

The simplestreamer tool uses incoming TCP ports 10000-10010. This needs to be added to the docs.

Document new meshconfig options

Add the following to meshconfig docs:

  • Document reference parameter
  • Document created_by parameter
  • Document configure_archives at test level
  • Document how multiple archives are handled
  • Document simplestream test
  • Document that guiagent can now do traceroute and ping tests

A couple regular testing notes to add

Our documentation should answer these:

  1. It seems this is a "push" architecture, i.e. the perfsonar node which
    made a measurement connects to the central MA to store the results there.

What happens if the central MA goes down for a while, or the network to
the central MA? Are the measurements lost, or are they buffered to be
delivered later? Does this have any effect on the timing of subsequent
measurements?

What if there are multiple MAs selected, but only one is down or
unreachable?

  1. If you have made manual changes to regular_testing.conf to point to a
    different MA (or multiple MAs), and you subsequently change test
    configurations through the GUI, does this leave your MA customisations
    alone?

Add a chapter about migrating toolkit from one host to another

Based on user request:
"I plan to migrate my perfsonarservers on a new hardware,
after the 3.5 version is installed and works on my old servers.

Is ther a list of what needs to be copied on the new servers?"

There probably just need to be a migration page added under "Managing the perfSONAR Toolkit” that highlights the important stuff.

Fix onpage search box

Add/fix Google widget to allow for searches to properly resolve using the onpage search box.

Extend documentation to define common test profiles

It would be good to extend current documentation to provide some common test profiles/templates i.e. toolkit settings (test duration, intervals, parameters, etc) for different measurement purposes to better reflect specific test traffic.

Debian/Ubuntu specific documentation

Most of the CLI related documentation we have needs to be reviewed in light of a Debian or Ubuntu installation. Let's track all the needed changes here.

Installation:

File locations:

Updates:

Central Server management:

Puppet for Debian?

We should double check the config files list:

Example of changes needed:

  • /sbin/service is /usr/sbin/service on Debian, but maybe we can refer to the command as service only.

List of services and cronjobs should be provided in documentation

From @arlake228 on March 14, 2015 12:52

Original issue 1071 created by arlake228 on 2015-02-03T15:10:01.000Z:

Describe the enhancement:

A user requested to add what exactly is the list of services that should be chkconfig'd on for a Latency box & for a Bandwidth box. Or maybe add a special note in 2.7.4 Special Upgrade Notes.

Example from user:
Not as in sec 2.4 via the web interface, but from cli:
chkconfig --list | sort | grep :on
List the perfsonar services for both Latency & Bandwidth please!
So if we find some obvious perfsonar ones list NOT on that list, we know there is some config problem.

Also there's only 1 perfsonar-installed crontab mentioned in doc (http://docs.perfsonar.net/multi_agent_config.html):
/etc/cron.d/cron-mesh_config_agent

But 4 others are installed (all had to be disabled while trying to clean up this poor troubled Latency box)
/etc/cron.hourly/logscraper.cron
/etc/cron.d/cron-clean_esmond_db
/etc/cron.d/cron-owamp_cleaner
/etc/cron.d/cron-service_watcher

Copied from original issue: perfsonar/project#1069

Add pS to other tools comparison

During the GEANT project review it was clear that some of the reviewers had difficulties in uderstanding (or missing knowledge) about main differences between perfSONAR and other performance monitoring tools like RIPE Atlas or SamKnows. It's proposed to add some kind of comparison table in FAQ in order to try highlighting main points in key areas.

delete hardware pages from www.perfsonar.net, and link to docs.perfsonar.net

user: "perfsonar-toolkit-servicewatcher" not "perfsonar-toolkit-service-watcher"

http://docs.perfsonar.net/release_candidates/4.0rc1/install_centos.html
Also under Step 2, its "perfsonar-toolkit-servicewatcher" not "perfsonar-toolkit-service-watcher" and the optional packages python install script simply doesnt exist at all as best as I can tell.

From: [email protected] [mailto:[email protected]] On Behalf Of Daniel Spisak
Sent: Mittwoch, 15. Februar 2017 04:39
To: [email protected]
Subject: [perfsonar-user] Questions about perfsonar 4

Document graphical installation for NetInstall and FullInstall

Both the full install and netinstall support a graphical installation method. We should document this with screenshots and text as the main Step-by-Step guide since it is the new default. The current content can be moved to a separate section in the same document on the text install.

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.