pandorafms/pandorafms

Pandora FMS is an open source monitoring application whose origin dates back to 2004. It integrates in the same application the monitoring of different infrastructure elements: networks, applications, servers, web, and other specific data sources such as logs, WMI, Netflow or SNMP traps.

With more than 50,000 customer installations across the five continents, Pandora FMS is an out-of-the-box monitoring solution: profitable and scalable, covering most infrastructure deployment options.

Pandora FMS gives you the agility to find and solve problems quickly, scaling them so they can be derived from any source, on-premise, multi cloud or both of them mixed. Now you have that capability across your entire IT stack and analytics to find any problem, even the ones that are hard to find.

Support

For community support you can visit our forums at https://pandorafms.com/community/

Pandora FMS has a "commercial" solution, with different features, oriented to companies that do not want to spend time using open source solutions, but closed packaged products, with periodic updates and professional support. Its name is Pandora FMS Enterprise, and you can find more information about it at https://pandorafms.com.

How to install Pandora FMS

Installing Pandora FMS from sources is a complex task. Please visit our wiki and follow all the steps described for a proper installation. https://pandorafms.com/manual/

What is Pandora FMS?

Pandora FMS is an open source monitoring application whose origin dates back to 2004. It integrates in the same application the monitoring of different infrastructure elements: networks, applications, servers, web, and other specific data sources such as logs, WMI, Netflow or SNMP traps.

It allows you to supervise systems and applications of all types, through remote monitoring or with software agents installed on the equipment to be monitored.

Pandora FMS monitors your hardware, software, your multilayer system and, of course, your operating system. Pandora FMS can detect if a network interface is down or the movement of the market value of any new NASDAQ technology. If desired, Pandora FMS can send an SMS message when your system or application fails or when the value of Tesla's stock drops below $180. Pandora FMS will adapt to your systems and requirements, because it has been designed to be open, modular, multiplatform and easy to customize.

Pandora FMS can be deployed over any OS, with specific agents for each platform. There are agents for Windows (2000, XP, 2003), GNU/Linux, Solaris, HP-UX, BSD, AIX, IPSO and OpenWRT. Pandora FMS not only gathers information through its agents, but it can also monitor any hardware system with TCP/IP connectivity, such as load balancing systems, routers, switches or printers, through SNMP and TCP/ICMP checks.

Pandora FMS is a monitoring tool that not only measures if a parameter is right or wrong. Pandora FMS can quantify the state (right or wrong), or store a value (numeric or alphanumeric) for months if necessary. Pandora FMS can measure performances, compare values among different systems and set alarms over thresholds. Pandora FMS works against a MySQL Database so it can generate reports, statistics, SLA and measure anything.

Main features

• Network monitoring
• Server monitoring (using agents for Windows, Linux, Mac, BSD and legacy Unix)
• SSH/WMI remote monitorin.
• Graphical reporting, based on SQL backend
• SLA, and ITIL KPI metrics on reporting
• Status & Performance monitoring
• GIS tracking and viewing
• Inventory management (Local and remote)
• Netflow support
• Centralized log collection.
• User defined visual console screens and Dashboards WYSIWYG
• Very high capacity (Thousands of devices)
• Multitenant, several levels of ACL management.

Screenshots

pandorafms's Issues

SQL error: Unknown column 'server_keepalive'

please, someone help me,

i got problem after update patch to 160516 like this,

SQL error: Unknown column 'server_keepalive' in 'where clause' ('SELECT COUNT(id_server) FROM tserver WHERE status = 1 AND keepalive > NOW() - INTERVAL server_keepalive*2 SECOND') in /var/www/pandora_console/include/db/mysql.php on line 76

i have do with repair table, but stil getting error in my pandora web console.

thanks

Agent Plugins don't work on 64 bit Windows

The standard conf files use %ProgramFiles% in the path to plugins but since the Pandora Agent is 32 bits it gets installed in Program Files (x86) so the plugins aren't found. The simplest solution is not to include a path at all since service's current directory is util.

Pandora subsystem inadvertenly down

Hi all,

we experienced an issue at one of our applications without Pandora noticing it, even though it was monitored. It turned out Pandora Server could not connect to the database:

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
2016-08-19 15:41:35 monep [V1] Launching 6.0SP2 160408
2016-08-19 15:41:35 monep [V1] Config options: Logfile at /var/log/pandora/pandora_server.log, Basepath is /etc/pandora/pandora_server.conf, Checksum is 0, Master is 1, SNMP Console is 0, Server Threshold at 5 sec, verbosity at 3, Alert Threshold at 60, ServerName is 'monep'
2016-08-19 15:41:35 monep [V1] [*] Pandora FMS Enterprise module not available.
2016-08-19 15:41:35 monep [V1] [E] 'monep': DBI connect('pandora:127.0.0.1:3306','pandora',...) failed: Can't connect to MySQL server on '127.0.0.1' (111) at /usr/lib/perl5/PandoraFMS/DB.pm line 119

2016-08-19 15:41:35 monep [V1] Pandora FMS Server 'monep' unhandled error.
2016-08-19 16:33:58 monep [V1] Launching 6.0SP2 160408
2016-08-19 16:33:58 monep [V1] Config options: Logfile at /var/log/pandora/pandora_server.log, Basepath is /etc/pandora/pandora_server.conf, Checksum is 0, Master is 1, SNMP Console is 0, Server Threshold at 5 sec, verbosity at 3, Alert Threshold at 60, ServerName is 'monep'
2016-08-19 16:33:58 monep [V1] [*] Pandora FMS Enterprise module not available.
2016-08-19 16:33:59 monep [V1] [E] 'monep': DBI connect('pandora:127.0.0.1:3306','pandora',...) failed: Can't connect to MySQL server on '127.0.0.1' (111) at /usr/lib/perl5/PandoraFMS/DB.pm line 119

2016-08-19 16:33:59 monep [V1] Pandora FMS Server 'monep' unhandled error.
2016-09-02 16:42:03 monep [V1] Launching 6.0SP2 160408
2016-09-02 16:42:03 monep [V1] Config options: Logfile at /var/log/pandora/pandora_server.log, Basepath is /etc/pandora/pandora_server.conf, Checksum is 0, Master is 1, SNMP Console is 0, Server Threshold at 5 sec, verbosity at 3, Alert Threshold at 60, ServerName is 'monep'
2016-09-02 16:42:03 monep [V1] [] Pandora FMS Enterprise module not available.
2016-09-02 16:42:04 monep [V1] [] Netflow daemon disabled.
2016-09-02 16:42:05 monep [V3] Warmup mode for unknown modules started.
2016-09-02 16:42:05 monep [V1] Server monep is the current master.
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

While this happened, the Pandora console showed the Events View, from where we didn't notice any error. Shouldn't we have received a much more obvious error message?

Thanks,
Sebastian-

No available data to show

I am testing pandora via docker and had

Is there anything I should do to show data and graphs of agent submitted modules (Module definition "Apprentice mode" is enabled):

Agent's log file is showing:
2017/01/21 03:36:13 - [log] - Running as user root
2017/01/21 03:36:13 - [setup] - server_ip is 10.99.5.7
2017/01/21 03:36:13 - [setup] - server_path is /var/spool/pandora/data_in
2017/01/21 03:36:13 - [setup] - temporal is /tmp
2017/01/21 03:36:13 - [setup] - logfile is /var/log/pandora/pandora_agent.log
2017/01/21 03:36:13 - [setup] - interval is 30
2017/01/21 03:36:13 - [setup] - debug is 0
2017/01/21 03:36:13 - [setup] - udp_server is 1
2017/01/21 03:36:13 - [setup] - udp_server_port is 41122
2017/01/21 03:36:13 - [setup] - udp_server_auth_address is 10.99.5.7
2017/01/21 03:36:13 - [setup] - agent_name is mns-thq-monitor-1
2017/01/21 03:36:13 - [setup] - group is Servers
2017/01/21 03:36:13 - [setup] - address is 10.99.7.3
2017/01/21 03:36:13 - [setup] - server_port is 41121
2017/01/21 03:36:13 - [setup] - transfer_mode is tentacle
2017/01/21 03:36:13 - [setup] - remote_config is 0
2017/01/21 03:36:13 - [setup] - xml_buffer is 1
2017/01/21 03:36:13 - [setup] - temporal_min_size is 1024
2017/01/21 03:36:13 - [log] - Thread is disabled.

Config file of agent is
pandora_agent.conf.txt

I'd like to specify all the module configuration on agent, and server to just swallow whatever I configure on agent. Is that possible?

[pandora_db.pl] mysql warning at purging old data

The following lines appeared at mysql log file;

140909 17:01:01 [Warning] Statement may not be safe to log in statement format. Statement: DELETE FROM tagente_datos_string WHERE utimestamp < 1402473134 AND utimestamp >= 1402473100 LIMIT 1000
140909 17:01:01 [Warning] Statement may not be safe to log in statement format. Statement: DELETE FROM tagente_datos_string WHERE utimestamp < 1402473202 AND utimestamp >= 1402473168 LIMIT 1000
140909 17:01:01 [Warning] Statement may not be safe to log in statement format. Statement: DELETE FROM tagente_datos_string WHERE utimestamp < 1402473270 AND utimestamp >= 1402473236 LIMIT 1000
140909 17:01:01 [Warning] Statement may not be safe to log in statement format. Statement: DELETE FROM tagente_datos_string WHERE utimestamp < 1402473508 AND utimestamp >= 1402473474 LIMIT 1000
140909 17:01:01 [Warning] Statement may not be safe to log in statement format. Statement: DELETE FROM tagente_datos_string WHERE utimestamp < 1402473542 AND utimestamp >= 1402473508 LIMIT 1000
140909 17:01:02 [Warning] Statement may not be safe to log in statement format. Statement: DELETE FROM tevento WHERE utimestamp < '1379145662' LIMIT 100
140909 17:01:02 [Warning] Statement may not be safe to log in statement format. Statement: DELETE FROM tevento WHERE utimestamp < '1379145662' LIMIT 100

I guess these would be written while db purging by pandora_db.pl.

Using binlog_format=mixed (instead of statement) can be a workaround,
but could I stop these warnings by correcting pandora_db.pl?
(adding order by ... or so?)

pandora-console docker image missing net-snmp-utils

...so SNMP wizard not working

Migration issue from 5.0 to 6.0

Hello Artica,

We are having an error on the migration from pandora version 5.0 to 6.0 when we run the alter pandoradb_migrate_5.1_to_6.0.mysql.sql:

ERROR 1072 (42000) at line 163: Key column 'id_rt' doesn't exist in table

As an anecdotic comment, when we perform the migration using the same sql script from a recent installed pandora version 5.1, and we havent that problem.

So the fact is:

Migration from clean pandora 5.0 to 5.1 = OK using pandoradb_migrate_5.0.x_to_5.1.mysql.sql
Migration from that pandora 5.1 to 6.0 = ERROR using pandoradb_migrate_5.1_to_6.0.mysql.sql

Migration from clean 5.1 to 6.0 = OK using the same pandoradb_migrate_5.1_to_6.0.mysql.sql

Could you please give us a hand with this? Do you have something that can fix this update problem?

Thank you for your time!

[Pandora Agent for Windows] util/getsnmp.bat doesn't see its installation directory.

If you install Pandora Agent for Windows at D: drive or so,
util/getsnmp.bat would fail to launch pandora_agent\util\snmpget.exe

This is because getsnmp.bat invoke snmpget.exe like;

"%ProgramFiles%\pandora_agent\util\snmpget" -M "%ProgramFiles%\pandora_agent\util\mibs" -Oveqtu -v 1 -c %1 %2 %3

SLA compliance not working on reports

Hello,
I am experiencing the following problem:

I used to be able to get correct values on the SLA compliance %. To be honest I don't know when it stopped working because I use them every 3 months.

With the latest version ( Pandora FMS v5.1SP3 - Build PC150624 ) I am not able to get it to work consistently.

Any ideas?

Feature Request: display event fired time on tables

Currently you are displaying the time since the event was fired, and only the time when you put the cursor over that time. It could be useful to add another column with that info or at least an option to change that behavior on settings.

<span class="datos_green"><span title="October 3, 2016, 11:22 pm" style="white-space:nowrap;">**11 horas**</span></span>

In a table like this I would like to see the time, to see the real downtime for that agent:

Wrong password parse form mobile login.

Noticed that passwords containing spaces throw an error (wrong user o password) in the mobile version of the login page '.../pandora_console/mobile/' while working correctly in the main login page.

Hidden comments in Events

I found a bug in event management. When I try to put a comment into one Event, the comment was stored in database, but when i try to reload in web page, the comment not appears.

This comment have a normal simbols like (, ), -> , new lines, etc... Not it´s a copy&paste procedure, I was written by hands in the comment box.

My pandora version is Pandora FMS v5.0SP3 - Revisión PC140128

_modulegraph_ alert not rendering

Hi,

I'm having an issue with sending a graph in an email alert. When configured, I receive a red x image.

I must be missing something obvious but I'm lost for ideas.

Group specified in agent's conf doesn't work when agent is created

If I specify a group in the agent's conf then when the first report happens the new agent is created with group Unknown. I've looked at the generated XML and the code on the server but I can't see anything obvious.

Topology Network Map uses wrong image for BSD systems

The file pandora_console/images/networkmap/so_bsd.png is actually the firewall icon. I've attached an resized version of the file pandora_console/images/os_icons/so_bsd.png to use instead.

Recon Task - Edit - Interval settings get lost / disabled task gets executed

When editing a Recon Task - the interval setting always get's set back to 10min, I have to change the field before saving to another value and then back..

If I saw it right, a not in list value get's lost while just opening the Recon Task

Did disable a task as well, it still got executed...

Alarms aren't restored automatically in Tree View

Hi all,

in Tree View, sometimes after a module changes to a normal state (from red, yellow or grey to green) color don't get updated automatically at the agent header (see screenshot) but only through a browser refresh.

Thanks,
Sebastian-

Mattermost moving to API version 4

Hey all!

First, hugely appreciate you sharing this project back with the world. Fantastic work!

I wanted to let you know about our plans moving to API v4:

To make the Mattermost API web service easier to use and to offer more powerful options for these integrations, Mattermost will be moving to a new API version soon. Highlights include:

• Fully documented API endpoints
• More in-depth access to server functionality
• Wider use of established HTTP verbs
• Consistent endpoint structures
• A new and improved Go driver

We plan to release API version 4 on March 16th, with Mattermost server 3.7. While the current API version 3 will be supported until September 16th, we recommend you begin using API version 4 soon after its release.

Contributing

API version 4 is an active and on-going project. If you're interested in helping contribute, please join our Mattermost community instance and the APIv4 channel.

We've prepared a contribution process for APIv4 and a progress tracker for new APIv4 enpoints to help you get started.

We're also open for suggestions on adding new API endpoints to help with your integration.

Change "UNKNOWN" color

Hi all,

sometimes its a bit hard to distinguish a green (normal) status from a grey (unknown) one. Is it possible to change the grey color to a different one?

Thanks,
Sebastian-

Configuring SSL in the tentacle server causes crash

If I follow the instructions for configuring the server to use SSL then it crashes when the first agent connects. I did some digging and it looks like a SIGSEGV in IO::Socket::SSL at line 1406 which is related to the password callback. Since I'm not using a password on the private key I commented out the SSL_passwd_cb parameter in start_ssl. That stopped the crash at line 1406 and added a new one at 1467 which is handling the SSL_verify_cb which isn't used.

This is on ver 5.1SP2 installed from the appliance CD.

Filtering by status in the Manage view

Hi all,

frequently we have to delete obsolete -grey- counters but to do so we have to search for them manually, on by one. Is there a way to filter grey counters from the "Manage agents" view?

Thanks,
Sebastian-

Network maps affected by Pandora 5.1 Upgrade

I recently upgraded from 5.0 to 5.1 and I have noticed a problem with network topology maps. The parent child node relationship for certain agents on my maps are not being followed. For example I have nine agents that have the Ethernet switch, and-ic3724pwr-b, as their parent; however, the topology network map for this location shows them linked directly to Pandora. So essentially, the parent entry on these agents is being ignored when the map is rendered in 5.1. I have attached an image of the map involved and I have confirmed that all of the agents have the correct parent data.

DB Status page:name of the field/COLUMN is missing in SQL ALTER statements

On the DB Status page (/pandora_console/index.php?sec=gdbman&sec2=extensions/db_status), when there is an error to the database the SQL query that can be used to fix it, does not contain the field name (should be inserted after the COLUMN statement).

Examples:

Oh and you might want to think of a single text box where you can find all SQL statements to fix the database at the end of the page. This way yoou don't have to do multiple copies, just 1 copy and paste to fix the database. Or maybe even better: add a button "perform fixes", that execute all INFORMATION SQL queries...

IPMI recon script doesn't work

There are a number of issues, the main one being it was never updated to use the new style custom fields although the plugin was updated.

I fixed this and some other issues in PR #15.

Dynamic counters

Hi all,

we wrote a plugin which connects to a database to be monitored, gets a list of its tablespaces and their usage and generates an XML for every tablespace.
When the database is unavailable an error XML should be generated. But because of this, a list of tablespaces cannot be built. Any suggestion or workaround? Maybe a generic counter could be added, which could act as a "flag". The alternative could be do nothing, so unknown alarms get fired.

Thanks,
Sebastian-

Reversed label in Template : Condition Type

In a Template, when "Condition Type" is "On Change" the check-box text next it is wrong. The text next to the box reads "Triggered when the value matches".

The informational text below "Condition Type" is actually correct: "The alert is triggered when the module's value changes".

DB Maintenance Task dies due to enterprise reference

The commit 64a595c needs to be cherry-picked from the develop branch to the Pandora_5.1 branch.

Pandora autorefresh sometimes doesnt work

Dear all,

We are using a pandora server 6.0. Its installed on a customer server, and we conect to that server trought the vpn. Sometimes and with no reason, the autorefresh option just stop to work, ande the events stop to refresh. Becouse of that we didnt see some of the alarms.

This may be a known issue? Its just something that is hhappening to us? We also tried to use a third party application for autorefresh but the result its the same.

I was looking the machine resources and the ram its ok (just 50% used), and cpu its sometimes beetween 60% to 90% . The machine have 7 pandora consoles opened in the same navigator. This could be a problem?

Any idea about what could be happening? What can we test?

Timezone_offset validation is incorrect

I have an agent in Dublin, one in Los Angles and my Pandora server is in New Zealand. This gives UTC TZs of 0, -8 and 13. i.e. Offsets of 13 and 21 respectively

In DataServer.pm, the following code needs to be updated
--- /tmp/DataServer.pm 2017-02-02 07:57:50.429557110 +1300
+++ DataServer.pm 2017-02-02 08:01:00.207712171 +1300
@@ -259,8 +259,8 @@
$data->{'interval'}, $data->{'os_version'}, $data->{'timezone_offset'},
$data->{'custom_id'}, $data->{'url_address'});

• Timezone offset must be an integer beween -12 and +12

• if (!defined($timezone_offset) || $timezone_offset !~ /[-+]?[0-9,11,12]/) {

• Timezone offset must be an integer beween -23 and +23

• if (!defined($timezone_offset) || $timezone_offset !~ /^[-+]?([0-9]|1[0-9]|2[0-3])$/) {
  $timezone_offset = 0;
  }

Fix the 'Report a Bug' link in the console

It would seem that SourceForge is no longer the place to report bugs/issues, if that is the case the link in the console should be updated to take one to GitHub instead.

Docker installation broken or not clear

After installing the latest stable version via:
http://wiki.pandorafms.com/index.php?title=Pandora:Documentation_en:Installing#Installation_under_Docker
curl -sSL http://pandorafms.org/getpandora | sh

the containers start but then nothing shows up if the server's ip is accessed - just the default Apache Centos page. The documentation is not clear what to do next:
http://wiki.pandorafms.com/index.php?title=Pandora:Documentation_en:Installing#Initial_Configuration_after_Installation
If I access the server like this:
http://<my_server_ip>/pandora_console/install.php

I get:
Not Found The requested URL /pandora_console/install.php was not found on this server.

This is by the way an awful first impression for the Pandorafms project - especially when one tries to evaluate it...

Auto refresh is 30 seconds on every page

I have just installed the VMware image and updated with yum. Every page has a 30 second auto refresh.
Pandora versions: v7.0NG.701 - Build PC170425 - MR 0

On most pages you can disable the auto refresh, although one that I found (create network map) seems to refresh instantly - a network map can not be created.

Thresholds on a plugin generating generic_data_inc

Hi all,

I have some plugins which generate XML output for modules of type generic_data. In these cases warning and critical thresholds are assined as plugin's parameters. The plugin calculates the module's value and compares it with the configured thresholds.
Now a new plugin generates generic_data_inc. In this case the plugin generates a "raw" output which is not the final data since the Pandora server calculates it as an increase. How should then the severity be calculated?

Regards,
Sebastian-

Issues with Passwords including $ and ! etc.

It seems like Pandora has issues while using Passwords including $ and ! signs for WMI etc...

As far as I could test it, the ! is the main issue.

Having several issues without a real explanation while using WMI walk and Passwords... always get instantly "unable to do wmi explorer" - a more detailed failure description would be nice...

Still, had some weird password issues all around... with $ and ! symbols in the password...

Cron Extension - Scheduled Jobs

Morning,

Running: Pandora FMS v6.0SP1 - Build PC160307 U606

I have an issue where it will not let me modify or delete any scheduled jobs. When clicking any option, I get redirected to a blank page. So currently, I have 2 jobs which are emailing fine daily, but I can't do anything with them.

Also if I go on custom reports & click send by email, nothing happens.

Thanks,
Rik

module_interval directive

Hi all,

I' trying to set an interval for a module generated by a plugin, an interval different from the agent's interval. I configured the module like this:

module_begin
module_plugin /INSWITCH/scripts/check/libexec/test
module_interval 3
module_end

And this is the code of the sample plugin (a Bash script):
echo "

Test counter
generic_data
NORMAL
$(date "+%s")
Test counter
<module_interval>3</module_interval>

"

The Pandora server receives the module's data at the interval defined for the agent and not for the module. It turns out that it doesn't work for Pandora FMS v5.0SP3 - Build PC140128, but does work for Pandora FMS v5.1SP3 - Build PC150624.

Thanks for your help,
Sebastian-

[Pandora Installation for Windows] Pandora FMS Server address - empty value breaks Nginx config

During Windows installation I have set an empty value in field Nginx Setup > Pandora FMS Server address. This has corrupted Nginx configuration so application cannot start. Please make this field mandatory (or set a default value) in Windows installer.

Steps to reproduce bug on Windows

• Download and execute Windows installer (pandorafms_server_6.0SP4_x86_64.exe)
• In panel Nginx Setup (see screenshot):
• Change value of field Pandora FMS Server address to empty (no value).
• Click on button Install to complete installation

Pandora FMS cannot start because Nginx is badly configured (see errors in C:\PandoraFMS\nginx\logs\error.log).

Workaround

Edit manually C:\PandoraFMS\nginx\conf\nginx.conf line 25:

        server_name  $hostname;

Variable name typo in pandora_mem plugin.

Hello.

I think I found a wrong variable name '$cachedwap' in pandora_mem plugin for unix, line 36.
The correct variable name probably should be '$cachedswap'.

https://github.com/pandorafms/pandorafms/blob/develop/pandora_agents/unix/plugins/pandora_mem

pandora-server docker run issue

root@builder:# docker run -p 41121:41121 --link pandora-mysql:mysql -d pandorafms/pandorafms-server:6
b0a10cb6348fef1b6ccdd1ac234ce9291a6eac7b93ce786aae803cf6fffbc061
root@builder:# docker ps -a |grep b0a10cb6348f
b0a10cb6348f pandorafms/pandorafms-server:6 "/entrypoint.sh" 43 seconds ago Exited (126) 41 seconds ago stoic_shaw
root@builder:# docker logs b0a10cb6348f
/entrypoint.sh: ./pandora_server_installer: /bin/sh: bad interpreter: Text file busy
root@builder:# docker start b0a10cb6348f
b0a10cb6348f
root@builder:# docker ps -a |grep b0a10cb6348f
b0a10cb6348f pandorafms/pandorafms-server:6 "/entrypoint.sh" About a minute ago Exited (9) 7 seconds ago stoic_shaw
root@builder:# docker logs b0a10cb6348f
/entrypoint.sh: ./pandora_server_installer: /bin/sh: bad interpreter: Text file busy
useradd: user 'pandora' already exists

root@builder:# cat /etc/*-release
DISTRIB_ID=Ubuntu
DISTRIB_RELEASE=16.04
DISTRIB_CODENAME=xenial
DISTRIB_DESCRIPTION="Ubuntu 16.04.1 LTS"
NAME="Ubuntu"
VERSION="16.04.1 LTS (Xenial Xerus)"
ID=ubuntu
ID_LIKE=debian
PRETTY_NAME="Ubuntu 16.04.1 LTS"
VERSION_ID="16.04"
HOME_URL="http://www.ubuntu.com/"
SUPPORT_URL="http://help.ubuntu.com/"
BUG_REPORT_URL="http://bugs.launchpad.net/ubuntu/"
UBUNTU_CODENAME=xenial
root@builder:# docker -v
Docker version 1.12.1, build 23cf638`

Group view display problem

Hi,
The group view display in Pandora 6 does not display the counts when the overall group bar is yellow, see attached screenshot.

Update mysql to mysqli

I have recently upgraded by servers to Fedora 25 which has PHP7 as the default package. mysql support has been removed from PHP7.

I found that by installing mysql_mysqli.inc.php in /var/www/html/pandora_console/include/db and adding a require_once in mysql.php, I got the console working with PHP7.

The correct solution would be to fix all the functions in mysql.php but I didn't have time to do that.

Secure Software Distribution?

The US government FedRAMP rules require all software to be cryptographically verified for integrity. The Pandora FMS repo does not appear to use a GPG key. Is there a GPG key available? Are there hashes of the repo files published on a secure site that we can use to verify the integrity of the software?

servidor http no puede escribir en el directorio de adjuntos

Acabo de instalar Pandora FMS en un servidor Centos 7. Tengo esta advertencia y no se como solucionarlo.
¿Qué permisos debo ponerle?
Desde ya, muchas gracias.

UNKNOWN status

When you create a template, to detect a module in an unknown state, this does not work so expected.

Server Versión 6.0SP2
Console Versión: Pandora FMS v6.0SP2 - Build PC160411

Sending some screenshots

XML_Buffer causes data to be sent out of order

If XML_Buffer is set to 1 then the agent sends the current data before processing the buffered data. This guarantees that generic_data_inc values will be considered invalid and cause log messages. In order to fix this the buffered ones would have to be sorted and sent in order before the new one.

This would also require changing the window's agent file names since they currently include a random number.

If you would like me to fix this I'll prepare a pull request.

pandora_db.pl: Use of uninitialized value in numeric gt (>) at /usr/share/pandora_server/util/pandora_db.pl line 96.

Cron was reporting this error every hour, on the db.pl script, please find below my fix suggestion.
(Pandora FMS DB Tool 7.0NG PS170324 Copyright (c) 2004-2015 Artica ST)

Problem:
/usr/bin/perl /usr/share/pandora_server/util/pandora_db.pl /etc/pandora/pandora_server.conf
...
Use of uninitialized value in numeric gt (>) at /usr/share/pandora_server/util/pandora_db.pl line 96.
..

Fix:
added a defined() check to the if clause.

        if (defined($conf->{'_inventory_purge'}) && $conf->{'_inventory_purge'} > 0) {
                if (enterprise_load (\%conf) != 0) {

Groupview

Since the update from 5.1SP4 to 6.0SP4 the Groupview doesn't seem to work properly anymore.
Once you click one of the Groups the Agent detail is working properly.

And just BTW, thank you for the great work. :)

API - get all_agents json option not working

The API documentation shows get all_agents has the option of return_type=json, and there is an an example call with json in both the code and documentation. However, the json option only returns csv-like output. My review of the code shows that
function api_get_all_agents in functions_api.php has the option of json output, but I find it only returns csv. I found that line 1077 has:
returnData('csv', $data, $separator);
whereas it should be:
returnData($return_type, $data, $separator);

Another needed change at line 961 is:
function api_get_all_agents($thrash1, $thrash2, $other, $return_type='csv') {
is replaced with:
function api_get_all_agents($thrash1, $thrash2, $other, $thrash3) {

Works!

Feautre Request: Agent VIEW - show all configured IP Address

Every now and then I get confused with the Agent-VIEW. I have several systems with multiple IP Addresses, but I don't see them directly in the Agent Overview - I only see one...

Would be nice to see all of them there...

Agent Communication on Windows Domain Controllers

Hi,

I've been setting up agents on our Windows 2008 R2 servers, and all are working fine with the exception of the agents deployed to domain controllers. I can install the agent on the DC's without issue, and the service also runs OK on these server. The DC's also appear in the Pandora console and look OK for about five minutes, but they then stop communicating and the console reports that they have passed their ETA.

The communication between the DC's and Pandora server is working fine, with no firewalls in the way, and as I've mentioned other servers (non DC's) on the same subnets work fine and the agents connect to the Pandora console and communicate properly.

Has anyone else seen this on Windows DC's? Is it a possibility that the DC is using a port that Pandora uses? It's also pretty weird that the DC's agent(s) register with the console initially, but then never re-communicate at the specified update time thereafter. I've checked the agent log and can't see anything and the agents on the DC's continue to run - they just don't connect to the server.

Cheers,

Kenny