joschi99/osiris

Osiris NMS

osiris's Issues

Upgrade to rancid 3.4.1

Upgrade rancid to version 3.4.1

Monitoring: WhatsApp notification channel

Create the possibility to use WhatsApp as a alternative notification channel for monitoring notifications using yowsup (https://github.com/tgalal/yowsup)

Osiris: change BI-S to PGUM GmbH

BI-S is now PGUM GmbH. Changing name, logos, disclaimer ecc on product

check_vmware_api.pl runtime health reports alert

HW situation:

• 2 x HP DL 380 (same hardware) in one vmware cluster (vsphere 1 + vsphere 2)
• 1 x HP DL 180 as standalone vmware server (vsphere 3)

SW situation:

• Osiris 2.1
• check_vmware_api.pl

vSphere situation:

• vCenter 5.5
• ESXi 5.1

The command is launched with the following parameters:

./check_vmware_api.pl -D vcenter55 -H vsphere1 -u osiris -p <password> -l runtime -s health -o listitems

On server vsphere 1 the command returns no alerts
On server vsphere 2 & vsphere3 the command returns 1 alert, but if you look the listitems every position will be ok and also and the vCenter will not report any error.

optimize centreon command configuration

Optimize the centreon command configuration:

• perl connector (performance)
• macros (centreon 2.7)

Upgrade GLPI 0.90.x

• GLPI 0.90.x
• FusionInventory 0.90+

Requirements:

• PHP 5.4.x (see #21)

Upgrade to rancid 3.3

Upgrade rancid to version 3.3 (changelog http://www.shrubbery.net/rancid/CHANGES)

Kaufgut: plugin missing for checking hardware on D-Link switch DES-3226S

The customer Kaufgut has a D-Link switch DES-3226S. The installed plugins are not working to monitor the hardware for this switch.

Twitter: "HTTP Error 403: Forbidden" when connect app to your Twitter account using register_app_to_account.py

For connecting the Twitter notifications to a specific Twitter Account you need to use the script register_app_to_account.py. Starting the script the following error is returned:

./register_app_to_account.py
Traceback (most recent call last):
File "./register_app_to_account.py", line 9, in
auth_url = auth.get_authorization_url()
File "/usr/lib/python2.6/site-packages/tweepy-2.1-py2.6.egg/tweepy/auth.py", line 96, in get_authorization_url
raise TweepError(e)
tweepy.error.TweepError: HTTP Error 403: Forbidden

It seems that the problem is that the request use HTTP and Twitter moved to HTTPS.

The script twitter_commandline.py for sending Twitter messages is working still again.

ELK update to version 2.x

Update the ELK-Stack to version 2.x

• Elasticsearch 2.x
• Logstash 2.x
• Kibana 4.3.x
• Shield 2.x
• Curator 3.4

[Centreon Plugins 20160902] - AKCP Sensorprobe: humidity warning/critical settings not working

Customer MIVOR reports that the humidity check on AKCP Sensorprobe 2 is not working very well.

./centreon_plugins.pl --plugin=hardware::sensors::akcp::snmp::plugin --mode=sensors --hostname=xxx.xxx.xxx.xxx --snmp-community=public --warning='temperature,.*,15:30' --warning='humidity,.*,30:60' --critical='temperature,.*,10:40' --critical='humidity,.*,20:70'

The plugin returns a lot of warning about the humidity, but the humidity values are inside the defined range. Same plugin with same parameters but empty humidity values are working correctly for temperature.
./centreon_plugins.pl --plugin=hardware::sensors::akcp::snmp::plugin --mode=sensors --hostname=xxx.xxx.xxx.xxx --snmp-community=public --warning='temperature,.*,15:30' --warning='humidity,.*,' --critical='temperature,.*,10:40' --critical='humidity,.*,'

Snare Agent for Windows: DYNAMIC Syslog Priority reverts to Emergency

With the latest Snare version 4.0.2.0 the syslog priority set to DYNAMIC switches back to "Emergency"

For further informations please visit also http://sourceforge.net/p/snare/bugs/39/

Upgrade Python 2.6 to a newer version

Osiris 2.2.0 and before has installed python 2.6 which is no longer supported and causes problems with certain library like urllib3 (need for twitter notification tweepy).
For this we will consider to upgrade python to a newer version like python 2.7 or 3.3

Implementing shield in elasticsearch

Implementing shield in elasticsearch to protect data and data access.

httpd failed to start/crash

On Osiris2 the httpd service (Apache 2.2) failed to start or crash.

Error on /var/log/http/error_log:

[Sun Sep 06 07:32:09 2015] [notice] suEXEC mechanism enabled (wrapper: /usr/sbin/suexec)
[Sun Sep 06 07:32:09 2015] [error] SSL Library Error: -8181 Certificate has expired
[Sun Sep 06 07:32:09 2015] [error] Unable to verify certificate 'Server-Cert'. Add "NSSEnforceValidCerts off" to nss.conf so the server can start until the problem can be resolved.

Elastic Beats integration

Integration of Elastic beats plattform

• Topbeat
• Packetbeat

[Kibana]: Map visualizations have stopped working due to MapQuest change July 11, 2016

MapQuest has change his access rules and the default Kibana map access is not more working.

Implement RELK-Stack

Implementing the full RELK-Stack (Logstash Shipper -> Redis -> Logstash Indexer -> Elasticsearch -> Kibana) in Osiris2. This feature extent the Logfile analysis to create different analysis and dashboards for different logfiles elaborating every logfile in a different way.

Netflow analysis using ELK

Configuring Logstash to parse NetFlow v5/v9 data streams and pass them on to Elasticsearch. Visualization with Dashboards Kibana

Nagios stops executing service checks

Customer Enghouse reports that since the change to wintertime the nagios daemon stops 2 times to execute service checks. It seems that the daemon stops at 0:00am

Systemdetails:

• Osiris 2.0
• Nagios 3.5.1
• ndoutils 1.5.2

dmesg.txt

[GLPI-Osiris2.2]: Usermanagement blank page

On Osiris2.1 and Osiris2.2 the usermanagement page shows only a blank page.

Osiris 2: Performance Problems -> CPU, Memory

Der Systemload von Osiris2 beim Kunden IT-Sonix (Enghouse) ist sehr hoch. Das System hat 16GB an RAM zur Verfügung, welche anscheinend auch verwendet werden und der CPU-Load steigt ständig an.

Laut Aussage des Kunden sind die darunterliegenden HD 10K-Platten und sollten genügend Performance liefern.

CPU Load:

VMWare Graphs:

[Osiris 2.2.0] Selenium 2.50: not working with firefox 45

Selenium 2.50 or older versions are not compatible with firefox 45.

IT Sonix - Centreon: Issue with ACL for the Graph widget

For every user without admin rights the graph widget show's the error "you're not allowed to reach this graph"

The error is documented on centreon forge under Bug #4614 (https://forge.centreon.com/issues/4614)

nxlog -> logstash: logswitch is behind 1 hour

On customer Kaufgut and on Osiris2.2 the logfiles that are coming from nxlog through logstash are written also to CIFS for digital signing. For every day the logs are in a specific directory. For all files coming from nxlog and logstash the logfile switch is not on 0.00am but on 1:00am.
This cause the problem that the digital signature (that will be applied on 0:05am) is not more valid for this files.

Remove Centreon-Syslog module

Centreon-Syslog module and functionality is superseded by the ELK solution and has to be removed:

• change rsyslog config (remove centreon-syslog DB)
• uninstall/deacivate centreon-syslog plugin
• remove centreon-syslog plugin
• modify backup script

Centreon Plugin Pack integration

Integration and full support of Centreon Plugin Pack in Osiris to support more checks and monitoring possibilities like:

• application: Apache, Asterisk, Elasticsearch, Github, Jenkins, Nginx, Pfsense, Tomcat, Varnish,...
• cloud: Docker
• database: Firebird, Informix, MS SQL, MySQL, Oracle, Postgres
• hardware: printers (rfc3805), UPS (Powerware, Mge, Standard), Sun Hardware, Cisco UCS, SensorIP, HP Proliant, HP Bladechassis, Dell Openmanage, Dell CMC, Raritan,...
• network: Aruba, Brocade, Bluecoat, Brocade, Checkpoint, Cisco AP/IronPort/ASA/Standard, Extreme, Fortigate, H3C, Hirschmann, HP Procurve, F5 BIG-IP, Juniper, PaloAlto, Redback, Riverbed, Ruggedcom, Stonesoft,...
• os: Linux (SNMP, NRPE), Freebsd (SNMP), AIX (SNMP), Solaris (SNMP)...
• storage: EMC Clariion, Netapp, HP MSA p2000, Dell EqualLogic, Qnap, Panzura, Synology...

note: install wsmancli for WSMAN checks

Upgrade Nagios Plugins

Upgrade Nagios Plugins from version 1.4.16 to version 2.1.1

IT Sonix - Centreon Servicegroup widget: duplicated services on servicegroup widget

https://forge.centreon.com/issues/4473
Problem is known and solution will be provided on new widget version 1.1.2 (installed version 1.1.0)

[database::mssql::plugin] How to connect to server

Using centreon plugins on MS SQL Server the connection give the following error:

perl centreon_plugins.pl --plugin=database::mssql::plugin --hostname=10.67.4.52 --port=1433 --mode=connection-time --username=* --password=*
UNKNOWN: Cannot connect: OpenClient message: LAYER = (0) ORIGIN = (0) SEVERITY = (78) NUMBER = (44)
Server , database 
Message String: Server name not found in configuration files.
OpenClient message: LAYER = (0) ORIGIN = (0) SEVERITY = (78) NUMBER = (45)
Server , database 
Message String: Unknown host machine name.

[Kaufgut] Osiris2.1 resize disk "/opt"

Customer ask to have 10GB more space on disk "/opt".

Centreon 2.7: Perl connector will not be exported on central poller

If I generate the configuration on the central poller the export will not find any connector and show's me this output:

Reading main configuration file '/usr/share/centreon/filesGeneration/nagiosCFG/1/nagiosCFG.DEBUG'.
Reading resource file '/usr/share/centreon//filesGeneration/nagiosCFG/1/resource.cfg'
Checking global event handlers...
Checking obsessive compulsive processor commands...

Checked 16 commands.
Checked 0 connectors.
Checked 1 contacts.
Checked 0 host dependencies.
Checked 0 host escalations.
Checked 1 host groups.
Checked 1 hosts.
Checked 0 service dependencies.
Checked 0 service escalations.
Checked 1 service groups.
Checked 29 services.
Checked 1 time periods.

If I export the configuration on the same system for the poller 1 then the connector will be found and exported:

Reading main configuration file '/usr/share/centreon/filesGeneration/nagiosCFG/2/nagiosCFG.DEBUG'.
Reading resource file '/usr/share/centreon//filesGeneration/nagiosCFG/2/resource.cfg'
Checking global event handlers...
Checking obsessive compulsive processor commands...

Checked 10 commands.
Checked 2 connectors.
Checked 1 contacts.
Checked 0 host dependencies.
Checked 0 host escalations.
Checked 1 host groups.
Checked 1 hosts.
Checked 0 service dependencies.
Checked 0 service escalations.
Checked 0 service groups.
Checked 8 services.
Checked 1 time periods.

Total Warnings: 0
Total Errors: 0

Issue open on centreon github (centreon/centreon#4045)

sflow analysis using ELK

Configuring Logstash to parse sflow data streams and pass them on to Elasticsearch. Visualization with Dashboards Kibana

Upgrade NagVis to version 1.8.x

Upgrade NagVis 1.7.10 to 1.8.x with centreon-nagvis-module support

Windows eventlog are not visible in kibana

The customer Kaufgut are not able to find the logs from the Windows Event Log in Kibana. The logs are send to Osiris2.1 and the are on the samba share, but in Elasticsearch the logs are not present.

Systems:

• Osiris 2.1
• RSyslog 8.13.0
• Elasticsearch 1.7.x
• Snare 4.0.2 for Windows

Selenium: basic authentication is not working

It is not possible to logon with Selenium on a website with basic authentication. The username and password is not filled in the combobox and firefox is not able to logon on the website with username and password via URL

nedi: missing definition files for HP Procurve 3500

Nedi is not able to scan correctly the HP Procurve 3500 because the definition files are missing.

nedi: missing definition files for HP Procurve 2530

Nedi is not able to scan correctly the HP Procurve 2530 because the definition files are missing.

Centreon 2.7 metaservice: Could not find broker connection info!

After upgrade to Centreon 2.7.0 the metaservices are not working and are in status "unknow" with status information "Could not find broker connection info!"

Centreon-Broker: RRD: could not connect to local socket '/var/rrdtool/rrdcached/rrdcached.sock: QLocalSocket::connectToServer: Invalid name

The Centreon-Broker are not able to generate the rrd graphs. In the logfile /var/log/centreon-broker/central-rrd-master.log are the following errors:

[1442406919] [0x1fc4d40] error:   RRD: could not connect to local socket '/var/rrdtool/rrdcached/rrdcached.sock: QLocalSocket::connectToServer: Invalid name

[GLPI]: error visualization My settings -> Personalization

On customer VI.P and also on Osiris 2.2.1 in GLPI the personalization page in "My settings" are displayed only in part and there is missing some information and there is no "save" button.
The apache error log shows the error:
PHP Fatal error: Allowed memory size of 67108864 bytes exhausted (tried to allocate 72 bytes) in /var/www/glpi/lib/tcpdf/fonts/cid0cs.php on line 15, referer: http://192.168.193.129/glpi/front/preference.php

PHP 5.4 update

Upgrade to php version 5.4.x on Scientific Linux 6.7

This is necessary to have more compatibility for future updates like GLPI 0.9 and others.

[GLPI - Fusioninventory]: Fusioninventory is changing the status

Customer MIVOR - Kurt Wellenzohn:

GLPI 0.90.3
Fusioninventory 0.90+1.2

It seems that Fusioninventory is changing the status from a existing object.

[rancid 3.4.1]: no matching cipher found

Rancid 3.4.1 in combination with a Dell N1524 switch reports on login using dllogin the error:

[rancid@osiris2` bin]$ ./dllogin 192.168.253.203
192.168.253.203
spawn ssh -c 3des -x -l admin 192.168.253.203
no matching cipher found: client 3des-cbc server aes128-ctr,aes192-ctr,aes256-ctr,arcfour,arcfour128,arcfour256,[email protected],

Error: Couldn't login: 192.168.253.203

Centreon 2.7.x: dashboard shows 100% undetermined

Since the upgrade to Centreon 2.7 the dashboard shows only 100% undetermined.
The cron eventReportBuilder runs every day on error:

2016-02-07 03:00:02 - Starting program...
2016-02-07 03:00:02 - Processing period: Sat Feb  6 00:00:00 2016 => Sun Feb  7 00:00:00 2016
Error: Can't call method "query" on an undefined value at /usr/share/perl5/vendor_perl/centreon/reporting/CentreonDownTime.pm line 96.

centreon-poller-centreon-engine-2.7.0: sed: -e expression #4, char 47: unterminated `s' command

Upgrading/installing/reinstalling on the poller the package centreon-poller-centreon-engine-2.7.0-7.el6.noarch returns error:

Running Transaction
  Installing : centreon-poller-centreon-engine-2.7.0-7.el6.noarch           1/1
sed: -e expression #4, char 47: unterminated `s' command
/var/tmp/rpm-tmp.KuzRwK: line 25: -e: command not found
  Verifying  : centreon-poller-centreon-engine-2.7.0-7.el6.noarch           1/1

Installed:
  centreon-poller-centreon-engine.noarch 0:2.7.0-7.el6

Complete!

Upgrade Centreon 2.7

Upgrade Centreon and components on Osiris 2.1 main and Poller to:

• Centreon 2.7.x
• Centreon Broker 2.11.x
• Centreon Engine 1.5.x
• Centreon Widgets
• Centreon Plugins

ITSonix: service widget shows incorrect datas

customer: ITSonix
service-widget: 1.3.2
centreon 2.5.x

The service widget has more issues:

• the number of items that are shown is not correct. The widget is confidured to show 10 items, but shows only 5 items
• in some cases there are missing the buttons for the other pages
• the total number of services on top of the item is not correct. The customer has 280 services configured, the widget report 484 items

[Centreon Plugins 20160902] - QNAP hardware - Disk with status normal cause critical error -> how we can change this to warning?

Customer MIVOR has a QNAP with 4 disks, where disk 3 has not a GOOD but a NORMAL (yellow) state. The
./centreon_plugins.pl --plugin=storage::qnap::snmp::plugin --mode=hardware --hostname=xxx.xxx.xxx.xxx --snmp-community=public --warning='temperature,.*,40' --critical='temperature,.*,50'

There is the parameter --threshold-overload to overload threshold values with default setting --threshold-overload='disk,CRITICAL,^(?!(ready)$)'

How is it possible to set the plugin that disk with status NORMAL is raising a WARNING?

nedi: missing definition files for HP Procurve 3500yl

Nedi is not able to scan correctly the HP Procurve 3500yl because the definition files are missing.