Sorry, I've forgotten to add users list.
---------- Forwarded message ---------- From: Игорь Коив koiv.igor@gmail.com Date: 2017-02-05 21:26 GMT+03:00 Subject: Re: Cisco SG500+NAV To: Morten Brekkevold morten.brekkevold@uninett.no
Good day, Morten!
Let me disturb you one more time. I have just installed Graphite and suppose, all got successful but then all the metrics disappeared. What could it be? Should I reinstall all NAV or can I detect the mistake?
BR, Igor Koiv
2017-02-02 9:56 GMT+03:00 Morten Brekkevold morten.brekkevold@uninett.no:
On Wed, 1 Feb 2017 22:27:23 +0300 Игорь Коив koiv.igor@gmail.com wrote:
Good Day!
Morten, there was no reason to delete the mailing list. I just didn't
want
to disturb all the comunity. Should I add it?
It's just that it's good form to always copy the list when the discussion started there - unless you wish to send information you want to keep out of the public. On the list there are potentially more responses to be had as well.
On the other hand, it's also not nice to bomb the list with screenshots, so it's good to post those in a private e-mail, or paste links to web versions.
Thank you for the response! It was not me who installed the NAV I read the manual but did nothing
about
the installation. I checked the way they had installed it and suppose it is "Quick start" version. Will it work properly? Should I install the full version?
Quick start is a full virtual appliance based on Debian 8 (Jessie). It should really work out of the box. If they installed the Debian package manually, there is more work to set things up.
Now daemon processes are run, the statistics is collected, suppose all
will
be excelent !!!
Graphite will be the next step.
Thanks a lot, sorry for the disturbance.
No problem, glad to hear it's working for you :-)
-- Morten Brekkevold UNINETT AS
On Mon, 6 Feb 2017 11:37:42 +0300 Игорь Коив koiv.igor@gmail.com wrote:
Let me disturb you one more time. I have just installed Graphite and suppose, all got successful but then all the metrics disappeared. What could it be? Should I reinstall all NAV or can I detect the mistake?
Your screenshots don't look like a Graphite issue to me. It looks rather like the Javascript part of the page failed to load. This sometimes happens when you upgrade NAV, as the browser keeps caching outdated resources. Usually, this is fixed by emptying the browser cache or force-reloading the page.
If emptying the browser cache doesn't work, then we need output from the browers web developer console to understand what's going on.
Good day! Thank you, Morten! All works properly after cleaning the browser cache.
One more thing: there are a lot of ports detected at SG-500. At 2960 all is ok. Should I disable them manually?
2017-02-07 11:11 GMT+03:00 Morten Brekkevold morten.brekkevold@uninett.no:
On Mon, 6 Feb 2017 11:37:42 +0300 Игорь Коив koiv.igor@gmail.com wrote:
Let me disturb you one more time. I have just installed Graphite and suppose, all got successful but then
all
the metrics disappeared. What could it be? Should I reinstall all NAV or can I detect the mistake?
Your screenshots don't look like a Graphite issue to me. It looks rather like the Javascript part of the page failed to load. This sometimes happens when you upgrade NAV, as the browser keeps caching outdated resources. Usually, this is fixed by emptying the browser cache or force-reloading the page.
If emptying the browser cache doesn't work, then we need output from the browers web developer console to understand what's going on.
-- Morten Brekkevold UNINETT AS
On Tue, 7 Feb 2017 23:24:29 +0300 Игорь Коив koiv.igor@gmail.com wrote:
Good day! Thank you, Morten! All works properly after cleaning the browser cache.
Glad to hear it :)
One more thing: there are a lot of ports detected at SG-500. At 2960 all is ok. Should I disable them manually?
Disabling the ports won't take them away. This is blown-up view is, unfortunately, a side effect of NAV not being able to get module information from the ports. If it did, they would at least by grouped by module.
The port reports might suit you better in this case.
Dear Morten! Let me disturb you again. I don't receive alert massages at all. Alertengine works, I'm admin with all permissions, Filter group "all filters", email is correct. We've disabled cables, but nothing received. what could it be?
2017-02-09 11:57 GMT+03:00 Morten Brekkevold morten.brekkevold@uninett.no:
On Tue, 7 Feb 2017 23:24:29 +0300 Игорь Коив koiv.igor@gmail.com wrote:
Good day! Thank you, Morten! All works properly after cleaning the browser cache.
Glad to hear it :)
One more thing: there are a lot of ports detected at SG-500. At 2960 all
is
ok. Should I disable them manually?
Disabling the ports won't take them away. This is blown-up view is, unfortunately, a side effect of NAV not being able to get module information from the ports. If it did, they would at least by grouped by module.
The port reports might suit you better in this case.
-- Morten Brekkevold UNINETT
On Fri, 17 Feb 2017 14:03:09 +0300 Игорь Коив koiv.igor@gmail.com wrote:
Dear Morten! Let me disturb you again.
I don't receive alert massages at all. Alertengine works, I'm admin with all permissions, Filter group "all filters", email is correct. We've disabled cables, but nothing received. what could it be?
You have, of course, verified that NAV actually shows alerts on the status page?
You say alertengine works, have you checked `alertengine.log`?
Do you have an MTA (like Postfix or Exim) installed on the NAV server, or did you configure `nav.conf` to use a different SMTP server than localhost?
Good day! Unfortunately, I have no alerts on the status page.
Yes, I've cheked it, it is empty. No data, but alertengine works.
Yes, I've configured `nav.conf` to use another SMTP server than localhost.
2017-02-21 10:27 GMT+03:00 Morten Brekkevold morten.brekkevold@uninett.no:
On Fri, 17 Feb 2017 14:03:09 +0300 Игорь Коив koiv.igor@gmail.com wrote:
Dear Morten! Let me disturb you again.
I don't receive alert massages at all. Alertengine works, I'm admin with all permissions, Filter group "all filters", email is correct. We've disabled cables, but nothing received. what could it be?
You have, of course, verified that NAV actually shows alerts on the status page?
You say alertengine works, have you checked `alertengine.log`?
Do you have an MTA (like Postfix or Exim) installed on the NAV server, or did you configure `nav.conf` to use a different SMTP server than localhost?
-- mvh Morten Brekkevold UNINETT
On Wed, 22 Feb 2017 00:07:46 +0300 Игорь Коив koiv.igor@gmail.com wrote:
Unfortunately, I have no alerts on the status page.
Then this isn't an issue with sending alerts at all, it's an issue with detecting something worth alerting about.
Can you please describe (in detail) what you are expecting NAV to alert you about, and what you are doing to provoke it?
Good day!
I've just expected to alert me about interfaces, switches, that are down or not reachable. And about users who has logged to the equipment. That is for the beginning.
We switched off the equipment, plugged off cables and logged to the switches. Physical disabling the interfaces and switches lasted for 10-20 min.
2017-02-22 13:16 GMT+03:00 Morten Brekkevold morten.brekkevold@uninett.no:
On Wed, 22 Feb 2017 00:07:46 +0300 Игорь Коив koiv.igor@gmail.com wrote:
Unfortunately, I have no alerts on the status page.
Then this isn't an issue with sending alerts at all, it's an issue with detecting something worth alerting about.
Can you please describe (in detail) what you are expecting NAV to alert you about, and what you are doing to provoke it?
-- Morten Brekkevold UNINETT
On Wed, 22 Feb 2017 13:30:20 +0300 Игорь Коив koiv.igor@gmail.com wrote:
I've just expected to alert me about interfaces,
By default, NAV will only alert you about uplinks/downlinks going down, and only if they represent loss of redundancy (in cases of no redundancy, you should get a boxDown alert for the no longer reachable devices).
switches, that are down or not reachable. And about users who has logged to the equipment. That is for the beginning.
We switched off the equipment, plugged off cables and logged to the switches. Physical disabling the interfaces and switches lasted for 10-20 min.
And the NAV backend processes are actually running (`nav status`)? What does `pping.log` say?
Yes, processes were running. pping.log worked, and detecteted that test interface was down and than up.
2017-02-22 14:26 GMT+03:00 Morten Brekkevold morten.brekkevold@uninett.no:
On Wed, 22 Feb 2017 13:30:20 +0300 Игорь Коив koiv.igor@gmail.com wrote:
I've just expected to alert me about interfaces,
By default, NAV will only alert you about uplinks/downlinks going down, and only if they represent loss of redundancy (in cases of no redundancy, you should get a boxDown alert for the no longer reachable devices).
switches, that are down or not reachable. And about users who has logged to the equipment. That is for the beginning.
We switched off the equipment, plugged off cables and logged to the switches. Physical disabling the interfaces and switches lasted for 10-20 min.
And the NAV backend processes are actually running (`nav status`)? What does `pping.log` say?
-- Morten Brekkevold UNINETT
On Wed, 22 Feb 2017 17:33:18 +0300 Игорь Коив koiv.igor@gmail.com wrote:
Yes, processes were running. pping.log worked, and detecteted that test interface was down and than up.
pping doesn't detect which interfaces are going up and down, it pings hosts.
Can you post an excerpt of `eventengine.log` from the same time period?
Also, please attach logs as text, not as images buried in openoffice documents.
Good day!
Sorry for the images, eventengine.log.1 in the attach.
2017-02-23 10:37 GMT+03:00 Morten Brekkevold morten.brekkevold@uninett.no:
On Wed, 22 Feb 2017 17:33:18 +0300 Игорь Коив koiv.igor@gmail.com wrote:
Yes, processes were running. pping.log worked, and detecteted that test interface was down and than up.
pping doesn't detect which interfaces are going up and down, it pings hosts.
Can you post an excerpt of `eventengine.log` from the same time period?
Also, please attach logs as text, not as images buried in openoffice documents.
-- Morten Brekkevold UNINETT
On Thu, 23 Feb 2017 13:46:46 +0300 Игорь Коив koiv.igor@gmail.com wrote:
2017-02-17 11:20:39,402 [INFO nav.eventengine.plugins.boxstate.BoxStateHandler] boxState start event for 192.168.4.201; warning in 60 seconds, declaring down in 240 seconds (if still unresolved)
[...]
2017-02-17 11:21:39,743 [INFO nav.eventengine.plugins.boxstate.BoxStateHandler] 192.168.4.201: Posting boxDownWarning alert
[...]
2017-02-17 11:24:40,148 [INFO nav.eventengine.plugins.boxstate.BoxStateHandler] 192.168.4.201: Posting boxDown alert
[...]
2017-02-17 11:36:15,123 [INFO nav.eventengine.plugins.boxstate.BoxStateHandler] 192.168.4.201 is back up
Everything is working as expected here. You should have been able to see an active alert about 192.168.4.201 being down in the Status tool of the web interface, between 11:24:40 and 11:36:15 on the 17th (unless you have filtered out these types of alerts from your status tool preferences).
So what does `alertengine.log` say in the same time period?
Sorry for the disturbing, I've missed check mark on "stateless" It shows all.
2017-02-23 14:51 GMT+03:00 Morten Brekkevold morten.brekkevold@uninett.no:
On Thu, 23 Feb 2017 13:46:46 +0300 Игорь Коив koiv.igor@gmail.com wrote:
2017-02-17 11:20:39,402 [INFO nav.eventengine.plugins.boxstate.BoxStateHandler]
boxState start event for 192.168.4.201; warning in 60 seconds, declaring down in 240 seconds (if still unresolved) [...]
2017-02-17 11:21:39,743 [INFO nav.eventengine.plugins.boxstate.BoxStateHandler]
192.168.4.201: Posting boxDownWarning alert [...]
2017-02-17 11:24:40,148 [INFO nav.eventengine.plugins.boxstate.BoxStateHandler]
192.168.4.201: Posting boxDown alert [...]
2017-02-17 11:36:15,123 [INFO nav.eventengine.plugins.boxstate.BoxStateHandler]
192.168.4.201 is back up
Everything is working as expected here. You should have been able to see an active alert about 192.168.4.201 being down in the Status tool of the web interface, between 11:24:40 and 11:36:15 on the 17th (unless you have filtered out these types of alerts from your status tool preferences).
So what does `alertengine.log` say in the same time period?
-- Morten Brekkevold UNINETT
On Thu, 23 Feb 2017 16:11:25 +0300 Игорь Коив koiv.igor@gmail.com wrote:
Sorry for the disturbing, I've missed check mark on "stateless" It shows all.
Hi again.
A checkmark on "Stateless" will show recent stateless alerts that were sent. A stateless boxDownWarning will be sent within 1 minute of a box stopping ping response. A stateful alert, which must be resolved again, will not be established until 4 minutes after the box stops responding to ping. When the stateful alert is resolved, it disappears from the status page.
Your alertengine.log indicates three different e-mails were sent to the same address in this time period. Likely a `boxDownWarning`, `boxDown` and later a `boxUp`.
If you didn't receive these e-mails, you need to check your e-mail servers or spam folder - the problem is outside of NAV.
Good day! Now I have the problem with Radius. I've installed freeradius than configure it with postgresql. But there is no path <path to freeradius>/etc/raddb/postgresql.conf
from this manual https://nav.uninett.no/doc/4.6/reference/radius-install.html?highlight=radiu...
so i can't do the next steps.
2017-02-23 17:50 GMT+03:00 Morten Brekkevold morten.brekkevold@uninett.no:
On Thu, 23 Feb 2017 16:11:25 +0300 Игорь Коив koiv.igor@gmail.com wrote:
Sorry for the disturbing, I've missed check mark on "stateless" It shows all.
Hi again.
A checkmark on "Stateless" will show recent stateless alerts that were sent. A stateless boxDownWarning will be sent within 1 minute of a box stopping ping response. A stateful alert, which must be resolved again, will not be established until 4 minutes after the box stops responding to ping. When the stateful alert is resolved, it disappears from the status page.
Your alertengine.log indicates three different e-mails were sent to the same address in this time period. Likely a `boxDownWarning`, `boxDown` and later a `boxUp`.
If you didn't receive these e-mails, you need to check your e-mail servers or spam folder - the problem is outside of NAV.
-- Morten Brekkevold UNINETT
On Thu, 16 Mar 2017 22:38:41 +0300 Игорь Коив koiv.igor@gmail.com wrote:
I've installed freeradius than configure it with postgresql. But there is no path <path to freeradius>/etc/raddb/postgresql.conf
from this manual https://nav.uninett.no/doc/4.6/reference/radius-install.html?highlight=radiu...
so i can't do the next steps.
I know next to nothing about FreeRADIUS. All I know is that this `postgresql.conf` file should be somewhere in your FreeRADIUS distribution. The same file is still referenced in the FreeRADIUS SQL-HOWTO wiki article:
https://wiki.freeradius.org/guide/SQL-HOWTO
koiv.igor@gmail.com said:
But there is no path <path to freeradius>/etc/raddb/postgresql.conf
You are supposed to create this file and make sure it is included from radiusd.conf. Or you can just edit the default sql.conf. Depending if which version of Freeradius you are using the file you need to modify might be located somwhere else.
From README.rst in freeradius-2.2.9:
" rlm_sql ~~~~~~~
The SQL configuration has been moved from ``sql.conf`` to ``mods-available/sql``. "
(You probably need to make a softlink or something from mods-enabled, but check the manual. - We use current software but our config "just works" and has been copied from older to newer software versions for ages...)
You want to use
driver = "rlm_sql_postgresql"
As for where the changes in SQL-queries goes I am really not sure. The file seems to be mods-config/sql/main/postgresql/queries.conf but the format has changed completely from when we made our config.
--Ingeborg
Thanks a lot! I hope all will be OK.
17 марта 2017 г. 13:36 пользователь "Ingeborg Hellemo" < ingeborg.hellemo@uit.no> написал:
koiv.igor@gmail.com said:
But there is no path <path to freeradius>/etc/raddb/postgresql.conf
You are supposed to create this file and make sure it is included from radiusd.conf. Or you can just edit the default sql.conf. Depending if which version of Freeradius you are using the file you need to modify might be located somwhere else.
From README.rst in freeradius-2.2.9: " rlm_sql
The SQL configuration has been moved from ``sql.conf`` to ``mods-available/sql``. " (You probably need to make a softlink or something from mods-enabled, but check the manual. - We use current software but our config "just works" and has been copied from older to newer software versions for ages...) You want to use driver = "rlm_sql_postgresql" As for where the changes in SQL-queries goes I am really not sure. The file seems to be mods-config/sql/main/postgresql/queries.conf but the format has changed completely from when we made our config. --Ingeborg -- Ingeborg Østrem Hellemo -- ingeborg.hellemo@uit.no Dep. of Information Technology --- Univ. of Tromsø
Dear developers! Could you help me again? What parameter should I use if I want to set threshold on load of the router port?
Is this correct:
*nav.devices.172_16_0_1.ports.Gi0_2.ifInOctets *sets traffic that goes in the router (bits/s) What threshold should be set? for example: *alert threshold >20000 and clear treshhold <19000 ?*
2017-03-17 14:58 GMT+03:00 Игорь Коив koiv.igor@gmail.com:
Thanks a lot! I hope all will be OK.
17 марта 2017 г. 13:36 пользователь "Ingeborg Hellemo" < ingeborg.hellemo@uit.no> написал:
koiv.igor@gmail.com said:
But there is no path <path to freeradius>/etc/raddb/postgresql.conf
You are supposed to create this file and make sure it is included from radiusd.conf. Or you can just edit the default sql.conf. Depending if which version of Freeradius you are using the file you need to modify might be located somwhere else.
From README.rst in freeradius-2.2.9: " rlm_sql
The SQL configuration has been moved from ``sql.conf`` to ``mods-available/sql``. " (You probably need to make a softlink or something from mods-enabled, but check the manual. - We use current software but our config "just works" and has been copied from older to newer software versions for ages...) You want to use driver = "rlm_sql_postgresql" As for where the changes in SQL-queries goes I am really not sure. The file seems to be mods-config/sql/main/postgresql/queries.conf but the format has changed completely from when we made our config. --Ingeborg -- Ingeborg Østrem Hellemo -- ingeborg.hellemo@uit.no Dep. of Information Technology --- Univ. of Tromsø
On Wed, 5 Apr 2017 22:04:33 +0300 Игорь Коив koiv.igor@gmail.com wrote:
Is this correct:
*nav.devices.172_16_0_1.ports.Gi0_2.ifInOctets *sets traffic that goes in the router (bits/s) What threshold should be set? for example: *alert threshold >20000 and clear treshhold <19000 ?*
If you want the threshold to be specifically 20000 bits/s for Gi0/2 on 172.16.0.1 (and the alert to be cleared when the traffic returns below 19000 bits/s), yes.
I might recommend that you use percentage values, though (especially if you are using a rule that will match multiple ports). The treshold monitor knows how to compare percentage values for octet counters using the nominal speed of the port (e.g. >90% and <89%)
Thanks a lot!
7 апр. 2017 г. 8:53 пользователь "Morten Brekkevold" < morten.brekkevold@uninett.no> написал:
On Wed, 5 Apr 2017 22:04:33 +0300 Игорь Коив koiv.igor@gmail.com wrote:
Is this correct:
*nav.devices.172_16_0_1.ports.Gi0_2.ifInOctets *sets traffic that
goes
in the router (bits/s) What threshold should be set? for example: *alert threshold >20000 and clear treshhold <19000 ?*
If you want the threshold to be specifically 20000 bits/s for Gi0/2 on 172.16.0.1 (and the alert to be cleared when the traffic returns below 19000 bits/s), yes.
I might recommend that you use percentage values, though (especially if you are using a rule that will match multiple ports). The treshold monitor knows how to compare percentage values for octet counters using the nominal speed of the port (e.g. >90% and <89%)
-- Morten Brekkevold UNINETT
-
Ingeborg Hellemo -
Morten Brekkevold -
Игорь Коив