I've setup a couple hosts running openbsd and note that uptime as reported in observium (I'm on the latest stable 13154) is showing incorrectly. Last reboot is showing as Jan 1 1970 as well.
Has anyone else seen this on OpenBSD?
Thanks,
Hi,
you have installed observium on OpenBSD or you monitor hosts with OpenBSD. (I didn’t really understand your notes)
Anyway.. make and attach (better in Discord pastebin channel) discovery and poller debugs for this host:
./discovery.php -d -h <device> ./poller.php -d -h <device>
Bryan Fields via observium wrote on 9.11.2023 11:58:
I've setup a couple hosts running openbsd and note that uptime as reported in observium (I'm on the latest stable 13154) is showing incorrectly. Last reboot is showing as Jan 1 1970 as well.
Has anyone else seen this on OpenBSD?
Thanks,
On 11/9/23 8:02 AM, Mike Stupalov via observium wrote:
Hi,
you have installed observium on OpenBSD or you monitor hosts with OpenBSD. (I didn’t really understand your notes)
I have observium installed on Linux, and added a couple new OpenBSD hosts as servers for it to monitor. On OpenBSD I am using the built in SNMPv3 server, which is as I understand it a different code base than typical net-snmp.
Anyway.. make and attach (better in Discord pastebin channel) discovery and poller debugs for this host:
Sorry, I don't have Discord, but I could jump on IRC if need be.
I've captured the output here: http://keekles.org/OpenBSD.txt
I didn't see an option to kill the escape codes for color though.
Thank you,
Hi,
Bryan Fields via observium wrote on 9.11.2023 20:03:
On 11/9/23 8:02 AM, Mike Stupalov via observium wrote:
Hi,
you have installed observium on OpenBSD or you monitor hosts with OpenBSD. (I didn’t really understand your notes)
I have observium installed on Linux, and added a couple new OpenBSD hosts as servers for it to monitor. On OpenBSD I am using the built in SNMPv3 server, which is as I understand it a different code base than typical net-snmp.
Anyway.. make and attach (better in Discord pastebin channel) discovery and poller debugs for this host:
Sorry, I don't have Discord, but I could jump on IRC if need be.
you can use IRC, we have gate between discord - irc (but I not remember which server) :)
I've captured the output here: http://keekles.org/OpenBSD.txt
I didn't see an option to kill the escape codes for color though.
That file ok, but this is not debug output.. you must run commands with -d flag as I written! With debug I can see required snmp outputs.. (or errors)
Thank you,
On 11/10/23 3:43 AM, Mike Stupalov wrote:
I've captured the output here:http://keekles.org/OpenBSD.txt
I didn't see an option to kill the escape codes for color though.
That file ok, but this is not debug output.. you must run commands with -d flag as I written! With debug I can see required snmp outputs.. (or errors)
Sorry, I know how frustrating that is, it was an oversight and I appreciate the help.
I've run both these commands: su observium -c '/opt/observium/discovery.php -d -h 52' >>/root/OpenBSD-debug.txt su observium -c '/opt/observium/poller.php -d -h 52' >>/root/OpenBSD-debug.txt
The output is here: http://keekles.org/OpenBSD-debug.txt
Thank you,
the escape codes can be a drag if you are trying to read in the browser, but if you download these (to a Linux host), and then view them with 'less -R <filename>', the colorization is rendered as you'd expect it to be, and may even help with reading.
Have you figured out the issue yet? I did not see (maybe missed) it get resolved here.
-C
On Fri, Nov 10, 2023 at 4:43 AM Bryan Fields via observium < observium@lists.observium.org> wrote:
On 11/10/23 3:43 AM, Mike Stupalov wrote:
I've captured the output here:http://keekles.org/OpenBSD.txt
I didn't see an option to kill the escape codes for color though.
That file ok, but this is not debug output.. you must run commands with -d flag as I written! With debug I can see required snmp outputs.. (or errors)
Sorry, I know how frustrating that is, it was an oversight and I appreciate the help.
I've run both these commands: su observium -c '/opt/observium/discovery.php -d -h 52'
/root/OpenBSD-debug.txt
su observium -c '/opt/observium/poller.php -d -h 52'
/root/OpenBSD-debug.txt
The output is here: http://keekles.org/OpenBSD-debug.txt
Thank you,
Bryan Fields
727-409-1194 - Voice http://bryanfields.net
observium mailing list -- observium@lists.observium.org To unsubscribe send an email to observium-leave@lists.observium.org
ok, couple disclaimers right up front: 1. I'm not yet running Observium - I've joined the list, but have yet to find the cycles to install it, and 2. I'm not an SNMP expert by any stretch
But while verifying that 'less -R' could indeed render your escape codes, it seem to me, unencumbered by knowledge, that you are missing mibs for specific items. It seems like it is returning data for basic stuff it understands, but it could simply be a confusion between net-snmp and ucd-snmp too, dunno. Hopefully you get this ironed out. Following, and interested, as I have some OpenBSD boxen myself...
-C
On Fri, Nov 10, 2023 at 4:43 AM Bryan Fields via observium < observium@lists.observium.org> wrote:
On 11/10/23 3:43 AM, Mike Stupalov wrote:
I've captured the output here:http://keekles.org/OpenBSD.txt
I didn't see an option to kill the escape codes for color though.
That file ok, but this is not debug output.. you must run commands with -d flag as I written! With debug I can see required snmp outputs.. (or errors)
Sorry, I know how frustrating that is, it was an oversight and I appreciate the help.
I've run both these commands: su observium -c '/opt/observium/discovery.php -d -h 52'
/root/OpenBSD-debug.txt
su observium -c '/opt/observium/poller.php -d -h 52'
/root/OpenBSD-debug.txt
The output is here: http://keekles.org/OpenBSD-debug.txt
Thank you,
Bryan Fields
727-409-1194 - Voice http://bryanfields.net
observium mailing list -- observium@lists.observium.org To unsubscribe send an email to observium-leave@lists.observium.org
Heh,
it's funny. Seems as this is an openbsd snmpd issue.
It's report 3 different Uptimes: 1. SNMPv2-MIB::sysUpTime.0 = 1:0:58:32.85 (Timeticks converted 1:0:58:32.85 -> 89912) 2. HOST-RESOURCES-MIB::hrSystemUptime.0 = 7:10:54:04.27 (Timeticks converted 7:10:54:04.27 -> 644044) 3. SNMP-FRAMEWORK-MIB::snmpEngineTime.0 = 1699608841
We always take 3 values and choose the maximum of them.
But seems as OpenBSD report just current unixtime in snmpEngineTime. This is incorrect.
I added workaround for this case in r13160.
Bryan Fields wrote on 10.11.2023 12:42:
On 11/10/23 3:43 AM, Mike Stupalov wrote:
I've captured the output here:http://keekles.org/OpenBSD.txt
I didn't see an option to kill the escape codes for color though.
That file ok, but this is not debug output.. you must run commands with -d flag as I written! With debug I can see required snmp outputs.. (or errors)
Sorry, I know how frustrating that is, it was an oversight and I appreciate the help.
I've run both these commands: su observium -c '/opt/observium/discovery.php -d -h 52'
/root/OpenBSD-debug.txt
su observium -c '/opt/observium/poller.php -d -h 52'
/root/OpenBSD-debug.txt
The output is here: http://keekles.org/OpenBSD-debug.txt
Thank you,
Knowing uptime is a security flaw, therefore this is a very very important OpenBSD security feature.
:D
adam.
Mike Stupalov via observium wrote on 13/11/2023 10:14:
Heh,
it's funny. Seems as this is an openbsd snmpd issue.
It's report 3 different Uptimes:
- SNMPv2-MIB::sysUpTime.0 = 1:0:58:32.85 (Timeticks converted
1:0:58:32.85 -> 89912) 2. HOST-RESOURCES-MIB::hrSystemUptime.0 = 7:10:54:04.27 (Timeticks converted 7:10:54:04.27 -> 644044) 3. SNMP-FRAMEWORK-MIB::snmpEngineTime.0 = 1699608841
We always take 3 values and choose the maximum of them.
But seems as OpenBSD report just current unixtime in snmpEngineTime. This is incorrect.
I added workaround for this case in r13160.
Bryan Fields wrote on 10.11.2023 12:42:
On 11/10/23 3:43 AM, Mike Stupalov wrote:
I've captured the output here:http://keekles.org/OpenBSD.txt
I didn't see an option to kill the escape codes for color though.
That file ok, but this is not debug output.. you must run commands with -d flag as I written! With debug I can see required snmp outputs.. (or errors)
Sorry, I know how frustrating that is, it was an oversight and I appreciate the help.
I've run both these commands: su observium -c '/opt/observium/discovery.php -d -h 52'
/root/OpenBSD-debug.txt
su observium -c '/opt/observium/poller.php -d -h 52'
/root/OpenBSD-debug.txt
The output is here: http://keekles.org/OpenBSD-debug.txt
Thank you,
-- Mike Stupalov, Discord channel: https://discord.gg/GjpNXKWm8W Observium Limited, https://observium.org
observium mailing list -- observium@lists.observium.org To unsubscribe send an email to observium-leave@lists.observium.org
On 11/13/23 5:14 AM, Mike Stupalov wrote:
Heh,
it's funny. Seems as this is an openbsd snmpd issue.
It's report 3 different Uptimes:
- SNMPv2-MIB::sysUpTime.0 = 1:0:58:32.85 (Timeticks converted
1:0:58:32.85 -> 89912) 2. HOST-RESOURCES-MIB::hrSystemUptime.0 = 7:10:54:04.27 (Timeticks converted 7:10:54:04.27 -> 644044) 3. SNMP-FRAMEWORK-MIB::snmpEngineTime.0 = 1699608841
We always take 3 values and choose the maximum of them.
But seems as OpenBSD report just current unixtime in snmpEngineTime. This is incorrect.
I suppose I should try reporting that to the maintainer. It's not Theo, so perhaps he'll listen :-)
I added workaround for this case in r13160.
Thanks Mike,
I switched to rolling to test it and can confirm the uptime is being reported correctly now.
participants (4)
-
Adam Armstrong -
Bryan Fields -
Christopher Barry -
Mike Stupalov