Nav-dev

nav-dev@lister.sikt.no

96 discussions

Patches for NAV-3.5.4
by Asbjørn Prøis 02 Nov '09

02 Nov '09

Attached is my patches for making NAV-3.5.4 compile'able with newer version of Cheetah. This is the same bug as this one : https://bugs.launchpad.net/nav/+bug/419750 -Asbjørn-

2 1

Zabbix?
by Vidar Faltinsen 27 Okt '09

27 Okt '09

nav-dev, In the network monitoring workshop in Belgrade that we (Morten, me, Arne) participated on last week... https://ow.feide.no/geantcampus:netw_monitoring_oct_2009#agenda ...I noticed that the University of Brno is using an open source product called Zabbix, implemented in Latvia. Have a look: http://www.zabbix.com/index.php Might be some inspiration there. They are definitely better on documentation than NAV (it seems). Look at the manual that is downloadable from http://www.zabbix.com/documentation.php We should have something like that for NAV. Challenge is to get someone to write this... Also AMRES (Serbian NREN) has implemented a solution called ICmynet. The first URL points to some presentations on the subject, I also have a user doc on that as well as a demo user for the netflow tool... - Vidar

1 0

Collecting VLAN data from Q-BRIDGE-MIB
by Morten Brekkevold 24 Sep '09

24 Sep '09

I've been looking at the current state of the vlan plugin to ipdevpoll, but it was quite unfinished and I decided to write something from scratch. I'm writing a plugin based solely on BRIDGE-MIB and Q-BRIDGE-MIB. Since Cisco has poor or no support for the latter, I'll later be writing a plugin to collect VLAN information from Cisco proprietary MIBs. My question is: Setting the vlan value of an interface record (previously swport) usually means the "configured" VLAN of that port. But there are possibly two ways to collect this from the Q-BRIDGE-MIB, and they have subtly different meanings. The options are basically ingress and egress VLANs. Q-BRIDGE holds a table of ports [1] with a VLAN (PVID) value for each. This is the VLAN assigned to any untagged ethernet frame coming in on that port. Then there is a table of active VLANs [2], each with a column listing egress ports that will send tagged or untagged frames from this VLAN. Which one should we choose? Should we look at both and do some sort of magic if they differ? [1] http://tools.cisco.com/Support/SNMP/do/BrowseOID.do?local=en&translate=Tran… [2] http://tools.cisco.com/Support/SNMP/do/BrowseOID.do?local=en&translate=Tran… -- mvh Morten Brekkevold UNINETT

2 1

NAV status meeting minutes 2009-05-05
by Morten Brekkevold 14 Mai '09

14 Mai '09

NAV status meeting minutes 2009-05-05 ===================================== Present: Morten, Magnus, Jørgen, Vidar Absent: Thomas, Kristian morten ------ * replacing psycopg1 with psycopg2 was a lot more work than first expected. lots of non-compliant DB-API usage, like: - dictfetch*() calls. - expectation of updateable dictionaries as result rows. - expectation of full pyformat template support in parameterized SQL queries. - django forces all strings returned by psycopg2 to be unicode objects, which was unexpected by lots of code. Then there are the apparent bugs in the psycopg2 versions available on Debian Etch and Lenny. * updated headers for webfront/subsystem and created an author list. Recommend everyone to go through code they're working on and do the same. * added logical BRIDGE-MIB instance scanning directly to getBoksMacs, per. blueprint "query-cisco-vlans-getboksmacs". Some research into what would be needed to support Q-BRIDGE-MIB from getBoksMacs as well - may be a very small task. magnus ------ * Working on refactor-auth blueprint. - broke login, fixed it, broke everything that uses sessions. - keep login, id and username in dictionary. - everything that accesses userdata as ForgetSQL object will fail atm. jørgen ------ * Studying for exams, but current task is smsd-backoff blueprint. etc --- * Assess whether we have enough new features to delay ipdevpoll to version 3.7 and release 3.6 on time. * One-day SNMP/gDD/ipdevpoll tutorial for summer interns in June, need to find a date that suits everyone. * We'll try daily 10 minute standing meetings each morning this summer to stay on top of things. -- mvh Morten Brekkevold UNINETT

1 0

Changelog RSS feed
by Morten Brekkevold 08 Mai '09

08 Mai '09

I was playing around with Yahoo Pipes, and created this feed: http://pipes.yahoo.com/pipes/pipe.run?_id=UMViT1Q73hGADMZeP27D0g&_render=rss This will return the last 100 changelog entries from any of the Mercurial branches published at http://metanav.uninett.no/hg/ . As a small bonus, the branch location (e.g. series/3.5.x) will be prepended to the item titles. Have a nice weekend! -- mvh Morten Brekkevold UNINETT

1 0

Authors list
by Morten Brekkevold 05 Mai '09

05 Mai '09

As previously decided, we will be removing author lists from individual source code files. This work is still in progress, and I just committed a few changesets to the default branch to update the headers of subsystem/webfront/. But, we do want to credit everyone who has contributed to NAV, so I've created a new top-level file for this: AUTHORS. I was in a verbose mode, and added a little text about the most significant contributors. Do we need this, or should we keep the list names only? Please have a look at http://metanav.uninett.no/hg/default/file/6b23e311c43a/AUTHORS and see if there are any glaring mistakes or omissions. Any comments about the structure/readability of the document would also be appreciated. -- mvh Morten Brekkevold UNINETT

2 1

NAV status meeting minutes 2009-04-21
by Morten Brekkevold 04 Mai '09

04 Mai '09

NAV status meeting minutes 2009-04-21 ===================================== Present: Morten, Magnus, Kristian Absent: Vidar, John-Magne, Thomas, Jørgen morten ------ * Working on ipdevpoll - Writing code for understanding MIBs, through SMI dumps from smilib. - Looking at alternatives for persisting collected data to database. Would like to use Django's ORM, but ORM's are in general ill suited for asynchronous work. * Working on implementing blueprint deprecate-psycopg1. * Looked at how forgetSQL can be removed from SeedDB. - forgetSQL is only used to read information. All writes are pure SQL. forgetSQL usage could easily be replaced by Django's ORM. - Complete rewrite in favor of Django's ORM and forms would probably simplify the entire subsystem. * Found that some people who experience frequent servicemon crashes are hit by a Linux kernel bug, which triggers an unhandled exception in calls to time.sleep(). magnus ------ * device history fixing * looked at ipdevpoll blueprints. * working on authentication-deprecate-forgetsql blueprint. - getting to know the codebase. kristian -------- * working mostly on main project (bachelor's degree) * fixed a couple of bugs (netmap, css) * just discovered a network explorer bug (session auth error thingy), investigating. jørgen (via email) ------------------ * A few report bugfixes from post-it notes. * Working on smsd-backoff blueprint. First attempt has been scrapped. --- I'll add that I know John-Magne was working on snmp trap support for WeatherGoose, and started research on mibs from UPS vendors APC and MGE. WeatherGoose support was completed, but needs testing and fine-tuning. -- mvh Morten Brekkevold UNINETT

1 0

Deprecating psycopg1
by Morten Brekkevold 22 Apr '09

22 Apr '09

Hi all! As mentioned on the IRC channel before easter week, I've been taking a shot at the deprecate-psycopg1 blueprint (https://blueprints.launchpad.net/nav/+spec/deprecate-psycopg1) Preliminary results can be found on this branch: http://metanav.uninett.no/hg/features/deprecate-psycopg1/ I would like to merge this code with the default branch as soon as possible. I would like it, though, if people who feel they have some sort of responsibility for code they've written would review the diffs. A lot of code was using the dictfetchall, dictfetchone and dictfetchmany call on cursors. These are non-standard, and are no longer supported in psycopg2. To avoid making too many code changes related to this, I've opted to use a psycopg2 extension which allows me to use customized cursor classes. The psycopg2 extensions already comes with a custom cursor class, which provides result rows that can be accessed both as tuples and dictionaries. To use the DictCursor class, I've had to manually specify the class as an argument to every call to connection.cursor(), where the code later expected to use the dictfetch methods on the cursor. All dictfetch calls have been reduced to regular fetch calls. The custom cursor change affects the following subsystems: * Arnold * deviceManagement * ipinfo * l2trace * lib-python (rrdPresenter) * maintenance * messages * Netmap * snmptrapd * statTools See this changeset to see how I may have mauled your code: http://metanav.uninett.no/hg/features/deprecate-psycopg1/rev/4cbc70349d55 On another note: Does all this code _really_ need to use dictionary result rows? I don't know the internals of psycopg2's DictCursor, but I seems to me that it could be a real memory-waste to pull thousands of table rows as dictionary instances instead of simple tuples. In many cases, I think looping over the results like this would suffice: cursor.execute("SELECT column1, column2, column3 FROM foo") for (column1, column2, column3) in cursor.fetchall(): pass # or actually do something -- mvh Morten Brekkevold UNINETT

1 3

NAV status meeting minutes 2009-03-23
by Morten Brekkevold 02 Apr '09

02 Apr '09

My apologies, I keep forgetting to post the meeting minutes to the list. On request, here are the minutes for last week's status meeting. Since next week is easter week, the next meeting will likely be April 21st, to stay on odd-numbered weeks. NAV status meeting minutes 2009-03-23 ===================================== morten ------ * released NAV 3.5.2 with two security fixes, coordinated with Werner for simultaneous release of the Debian package. * fixed a couple of ipdevinfo bugs (rrd-availability & downtime for stateless alerts) * Fixed Netmap UTF-8 bug for Windows users. Kristian comments that the old convoluted code was an attempted workaround for some older java plugins. My take is that we should recommend everyone always use the latest available version. * using virtualbox on my workstation for deployment tests, old sdrift machine decommissioned. * new virtual machine, elixis, for ipdevpoll development. * "apache idle in transactions"-problem on client installations fixed. Was related to Apache site config overriding NAV's .htaccess file. * consolidated interface table looks complete. Schema upgrade statements written and tested. Should have a short meeting about this when Vidar gets back fro Lisbon. jørgen ------ * working on smsd backoff mechanism magnus ------ * wrote a Django base template to have SQL query debug info on all pages in Django subsystems. * Hacking on error message logging in IP Device history. Found a quickselect defect: Cannot explicitly set a default selection, e.g. when form needs to be redisplayed because of errors in form input. thomas ------ * alertengine bugfixing. discussion ---------- Questions about priorities: Magnus: Priority list: 1=finish ipdevhist, 2=ipdevpoll tasks, if any are assigned, 3=refactor-auth blueprint Jørgen: Priority list: 1=smsd backoff, 2=ipdevinfo-related bugs and blueprints, 3=report-sql-parse-refactor -- mvh Morten Brekkevold UNINETT

1 0

Altered db.conf semantics
by Morten Brekkevold 31 Mar '09

31 Mar '09

Hi everyone, Just thought I should give a heads up that I have just implemented the following blueprint: https://blueprints.launchpad.net/nav/+spec/reduce-dbconf-complexity The changes have been merged to the default branch, see http://metanav.uninett.no/hg/default/log/83c18db13fdc . For those using the APIs provided with NAV to get DB connection parameters there should be no problems. But if you have some hackish code that parses or uses options from db.conf outside of the supported APIs, be aware that a a lot of options have now disappeared from the default db.conf. Of the script_* and db_* parameters, only "script_default" and "db_nav" are now required. The APIs will fall back to these two options when the wanted script_* and db_* options are missing. You can still connect as a different user if needed (e.g. for debugging purposes) by adding the needed options as before. -- mvh Morten Brekkevold UNINETT

1 0

← Newer
1
2
3
4
5
6
7
8
9
10
Older →

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

Nav-dev