Gaps in some graphs, not others

[TheWitness]

Please send via e-mail. Also, send some of the other LOW log information with it for several polling cycles.

TheWitness

[TheWitness]

I am getting behind myself. This is a known problem with the current version. I have two issues to resolve this weekend. However, the SVN version of Cactid already remedies this issue. Also, per my e-mail, you should increase your snmp timeout.

TheWitness

[raccoon]

Hey all, just wanted to point out that I just resolved a similar gap problem in my graphs by checking my crontabs - as stated in a post above.

I recently switched to a manual install of the latest cacti version, instead of the Debian package I had installed before (one nasty db bug there), and forgot to remove the old cron job before adding the new one

Check this first before losing your temper!

[kwabbernoot]

I am getting behind myself. This is a known problem with the current version. I have two issues to resolve this weekend. However, the SVN version of Cactid already remedies this issue. Also, per my e-mail, you should increase your snmp timeout.

TheWitness

Hi,

I am having a similar problem,

I donwloaded the CVS version using the svn command syntax included on the site, but this cactid tells me it is version 0.9 and it has a compilation problem.

Is there a patched cactid version 0.8.6h available ??

With kind regards,
Kwabbernoot

[kwabbernoot]

I am getting behind myself. This is a known problem with the current version. I have two issues to resolve this weekend. However, the SVN version of Cactid already remedies this issue. Also, per my e-mail, you should increase your snmp timeout.

TheWitness

Hi,

I am having a similar problem,

I donwloaded the CVS version using the svn command syntax included on the site, but this cactid tells me it is version 0.9 and it has a compilation problem.

Is there a patched cactid version 0.8.6h available ??

With kind regards,
Kwabbernoot

Sorry,

My mistake, I posted too quickly

Regards,
Kwabbernoot

[fmangeant]

Hi Larry

is it safe to use the SVN version of cactid ? Or should we wait until 0.8.6h release ?

Thanks in advance for your answer.

[santiagosoares]

Hello,

I have a similar problem, here, The last hosts added have some gaps in the graphs. Did somebody find a solution??

[ssozonoff]

Guys,

I am seeing this too

Code: Select all

07/03/2006 08:05:44 PM - CMDPHP: Poller[0] ERROR: SQL Assoc Failed "select   snmp_query_graph_rrd.snmp_field_name,   data_template_rrd.id as data_template_rrd_id   from (snmp_query_graph_rrd,data_template_rrd)   where snm
p_query_graph_rrd.data_template_rrd_id=data_template_rrd.local_data_template_rrd_id   and snmp_query_graph_rrd.snmp_query_graph_id=   and snmp_query_graph_rrd.data_template_id=41   and data_template_rrd.local_data_id=57"
07/03/2006 08:05:44 PM - CMDPHP: Poller[0] ERROR: SQL Assoc Failed "select   snmp_query_graph_rrd.snmp_field_name,   data_template_rrd.id as data_template_rrd_id   from (snmp_query_graph_rrd,data_template_rrd)   where snm
p_query_graph_rrd.data_template_rrd_id=data_template_rrd.local_data_template_rrd_id   and snmp_query_graph_rrd.snmp_query_graph_id=   and snmp_query_graph_rrd.data_template_id=41   and data_template_rrd.local_data_id=69"
07/03/2006 08:05:44 PM - CMDPHP: Poller[0] ERROR: SQL Assoc Failed "select   snmp_query_graph_rrd.snmp_field_name,   data_template_rrd.id as data_template_rrd_id   from (snmp_query_graph_rrd,data_template_rrd)   where snm
p_query_graph_rrd.data_template_rrd_id=data_template_rrd.local_data_template_rrd_id   and snmp_query_graph_rrd.snmp_query_graph_id=   and snmp_query_graph_rrd.data_template_id=41   and data_template_rrd.local_data_id=73"
07/03/2006 08:05:44 PM - CMDPHP: Poller[0] ERROR: SQL Assoc Failed "select   snmp_query_graph_rrd.snmp_field_name,   data_template_rrd.id as data_template_rrd_id   from (snmp_query_graph_rrd,data_template_rrd)   where snm
p_query_graph_rrd.data_template_rrd_id=data_template_rrd.local_data_template_rrd_id   and snmp_query_graph_rrd.snmp_query_graph_id=   and snmp_query_graph_rrd.data_template_id=41   and data_template_rrd.local_data_id=80"
07/03/2006 08:05:44 PM - CMDPHP: Poller[0] ERROR: SQL Assoc Failed "select   snmp_query_graph_rrd.snmp_field_name,   data_template_rrd.id as data_template_rrd_id   from (snmp_query_graph_rrd,data_template_rrd)   where snm
p_query_graph_rrd.data_template_rrd_id=data_template_rrd.local_data_template_rrd_id   and snmp_query_graph_rrd.snmp_query_graph_id=   and snmp_query_graph_rrd.data_template_id=41   and data_template_rrd.local_data_id=87"

No idea why it has started to show itself. The only thing we have done is added a load more devices and graphs....

Serge

[kwabbernoot]

I upgraded my cactid version with the version I pulled from the SVN Cacti 086 branch.

My Cacti Production enviroments both use this patched cactid version and I must say that I do not notice any problems.
Errors logs that were generated by the unpatched cactid version did disappear after the upgrade.

However my gaps in the Barracuda devices did not disappear but this is due to another cause.
These ******* barracuda developers find it interesting to restart the SNMP agent whenever they feel like it and thus Cacti re-caches the device every time the agent restarts.

More info about this in the following post:
http://forums.cacti.net/viewtopic.php?t ... highlight=

Regards,
Kwabbernoot

[AdrianW]

I upgraded my cactid version with the version I pulled from the SVN Cacti 086 branch.

My Cacti Production enviroments both use this patched cactid version and I must say that I do not notice any problems.
Errors logs that were generated by the unpatched cactid version did disappear after the upgrade.

However my gaps in the Barracuda devices did not disappear but this is due to another cause.
These ******* barracuda developers find it interesting to restart the SNMP agent whenever they feel like it and thus Cacti re-caches the device every time the agent restarts.

More info about this in the following post:
http://forums.cacti.net/viewtopic.php?t ... highlight=

Kwabbernoot, Hello. I am experiencing the same issue in which gaps are being produced with a couple Barracuda devices trying to monitor the interface traffic and CPU usage on them (other values are graphed/polled without issue, including - Load Averages, Memory Usage, and Email Queues). I reviewed the other post you referenced as well. I am working with an unpatched, "manual" installation of Cacti 0.8.6h.

Do you have any specific evidence (i.e. documentation, forum posts) of the Barracuda devices restarting the SNMP agent as you suggest? Or is this just based on your (seemingly great) experience? Finally, I assume you have not yet found a working solution for these Barracuda devices. However, if you have found a solution, would you be willing to share?

Thank you.

[theitch]

I am having the same problem. The poller (via cactid) looks like it is trying to run an update against an rrd file. You can see in 1), cacti attempts to update the rrd file in question. However, in 2), the file doesn't reflect the last update.

What's additionally aggrivating is that it is somewhat sporadic for *some* graphs. Over the weekend, some data was written to the proper rrd files, though I didn't touch the system. Though most graphs don't work.

Lastly, I've tried clearing the poller cache numerous times. Still no luck.
Any suggestions on what to try next are appreciated!

Robb

VERSIONS:
php-4.3.10
mysql-3.23.58
cacti-0.8.6h
rrdtool-1.0.50
perl-5.8.3

EXAMPLE DATA FOR ONE DATASOURCE/HOST/GRAPH (longshot traffic in and out):

1) DEBUG OUTPUT FROM poller.php
07/25/2006 05:41:07 PM - POLLER: Poller[0] CACTI2RRD: /usr/bin/rrdtool update /opt/www/html/cacti-0.8.6h/rra/lifeguard_traffic_in_153.rrd --template traff
ic_out:traffic_in 1153874449:3150336201:1412683133


2) COMMAND: rrdtool info lifeguard_traffic_in_153.rrd
OUTPUT:
filename = "lifeguard_traffic_in_153.rrd"
rrd_version = "0001"
step = 300
last_update = 1153426138
ds[traffic_in].type = "COUNTER"
ds[traffic_in].minimal_heartbeat = 600
ds[traffic_in].min = 0.0000000000e+00
ds[traffic_in].max = 1.0000000000e+10
ds[traffic_in].last_ds = "2908092532"
ds[traffic_in].value = 1.4690719798e+08
ds[traffic_in].unknown_sec = 0
ds[traffic_out].type = "COUNTER"
ds[traffic_out].minimal_heartbeat = 600
ds[traffic_out].min = 0.0000000000e+00
ds[traffic_out].max = 1.0000000000e+10
ds[traffic_out].last_ds = "3160951814"
ds[traffic_out].value = 2.0579142504e+08
ds[traffic_out].unknown_sec = 0
rra[0].cf = "AVERAGE"
rra[0].rows = 600
rra[0].pdp_per_row = 1
rra[0].xff = 5.0000000000e-01
rra[0].cdp_prep[0].value = NaN
rra[0].cdp_prep[0].unknown_datapoints = 0
rra[0].cdp_prep[1].value = NaN
rra[0].cdp_prep[1].unknown_datapoints = 0
rra[1].cf = "AVERAGE"
rra[1].rows = 700
rra[1].pdp_per_row = 6
rra[1].xff = 5.0000000000e-01
rra[1].cdp_prep[0].value = 6.2039270093e+05
rra[1].cdp_prep[0].unknown_datapoints = 0
rra[1].cdp_prep[1].value = 8.6133458180e+05
rra[1].cdp_prep[1].unknown_datapoints = 0

[TheWitness]

How many cron poller tasks do you have running? Did you install from a distribution or from source? Run the poller from a shell and look for error messages.

TheWitness

[theitch]

Thanks for responding.
cron poller runs every 5 minutes
Cacti install is from an RPM. I built the RPM from this SRPM: http://www.cacti.net/downloads/packages ... -1.src.rpm

See below for the following actions & output:
1) run poller.php redirecting stdout and stderr to files.
2) output of stderr from poller.php
3) last two lines of stdout
4) rrdtool info for the last rrd file updated from the execution of the poller.

[cactiuser@host cacti]$ php poller.php >/tmp/cacti.poller.stdout.log 2>/tmp/cacti.poller.stderr.log
[cactiuser@host cacti]$ cat /tmp/cacti.poller.stderr.log
snmpwalk: Error building ASN.1 representation
snmpwalk: Error building ASN.1 representation
snmpwalk: Error building ASN.1 representation
[cactiuser@host cacti]$ tail -2 /tmp/cacti.poller.stdout.log
07/26/2006 11:11:50 AM - POLLER: Poller[0] CACTI2RRD: /usr/bin/rrdtool update /opt/www/html/cacti-0.8.6h/rra/madrox_proc_791.rrd --template proc 1153937491:161
07/26/2006 11:11:50 AM - SYSTEM STATS: Time:18.1764 Method:cactid Processes:1 Threads:50 Hosts:59 HostsPerProcess:59 DataSources:1407 RRDsProcessed:850
[cactiuser@cerebro cacti]$ rrdtool info /opt/www/html/cacti-0.8.6h/rra/madrox_proc_791.rrd
filename = "/opt/www/html/cacti-0.8.6h/rra/madrox_proc_791.rrd"
rrd_version = "0001"
step = 300
last_update = 1153426113
ds[proc].type = "GAUGE"
ds[proc].minimal_heartbeat = 600
ds[proc].min = 0.0000000000e+00
ds[proc].max = 1.0000000000e+03
ds[proc].last_ds = "UNKN"
ds[proc].value = 1.1715000000e+04
ds[proc].unknown_sec = 0
rra[0].cf = "AVERAGE"
rra[0].rows = 600
rra[0].pdp_per_row = 1
rra[0].xff = 5.0000000000e-01
rra[0].cdp_prep[0].value = NaN
rra[0].cdp_prep[0].unknown_datapoints = 0
rra[1].cf = "AVERAGE"
rra[1].rows = 700
rra[1].pdp_per_row = 6
rra[1].xff = 5.0000000000e-01
rra[1].cdp_prep[0].value = 6.5966666667e+01
rra[1].cdp_prep[0].unknown_datapoints = 0
--OUTPUT TRUNCATED--

[bchecketts]

I also had this same problem after moving cacti to a new server. I tried to fix it for over a month whenever I had some available time. I had added tons of extra debugging to the php code and could never figure it out.

Finally I disabled the cron job (the one that I knew of) and was tailing the cacti log to find it still running somehow! I found out that the second mysterious job was running from the package that I had installed from rpmforge in /etc/cron.d/cacti


Hope this helps somebody.
Brandon Checketts
www.gamepricewatcher.com