choppy graphs

MrJake · Post by **MrJake** » Sat Sep 27, 2008 11:05 am

I am having choppy and incomplete graphs. Im using spine to poll. I set it to 1 process and 10 threads and 1 oid per poll. here is the stats of the poll:
Server 1
09/27/2008 11:55:06 AM - SYSTEM STATS: Time:5.0681 Method:spine Processes:1 Threads:10 Hosts:10 HostsPerProcess:10 DataSources:3875 RRDsProcessed:1482

Server 2
09/27/2008 11:55:13 AM - SYSTEM STATS: Time:11.0708 Method:spine Processes:1 Threads:5 Hosts:11 HostsPerProcess:11 DataSources:3040 RRDsProcessed:1269

Server 1 is a vmr esx server this server has 2 cores at 2ghz and 1 gig of ram

server 2 is an old compaq proliant something running 2 pentium 3 1ghz with 1 gig of ram

Serevr 1 has a true gig feed to most of these device and server 2 has a 100mb feed.

Both have choppy graphs. what is most puzzling is i have 1 device which is a cisco 3750 that its graph are never choppy. but all other devices most of the graphs a like the ones includes.

So what I did last night is put server 1 in motion to see if it was server 2 having hardware issues and i still get choppy graphs on that one. the choppy graphs are from a ubr10012 running at 6% cpu. so regardless if its a beast or some lagged out cmts i get choppy graphs on most my polls.

Any clues?

Post by **TheWitness** » Mon Oct 06, 2008 8:10 pm

Check for two cron entries first. Then, look to see for any errors, or timeouts in the Cacti log. Post your findings.

TheWitness

MrJake · Post by **MrJake** » Wed Oct 08, 2008 4:08 am

hiddenhn:~# crontab -l -u cactiuser
0,5,10,15,20,25,30,35,40,45,50,55 * * * * php /var/www/cacti/poller.php > /dev/null 2>&1
hiddenhn:~#

logs
10/08/2008 04:55:21 AM - SYSTEM STATS: Time:19.5778 Method:cmd.php Processes:3 Threads:N/A Hosts:11 HostsPerProcess:4 DataSources:3040 RRDsProcessed:1269
10/08/2008 04:50:22 AM - SYSTEM STATS: Time:19.7052 Method:cmd.php Processes:3 Threads:N/A Hosts:11 HostsPerProcess:4 DataSources:3040 RRDsProcessed:1269
10/08/2008 04:45:21 AM - SYSTEM STATS: Time:19.3111 Method:cmd.php Processes:3 Threads:N/A Hosts:11 HostsPerProcess:4 DataSources:3040 RRDsProcessed:1269

Sysinfo:

Cacti Version - 0.8.7a
Plugin Architecture - 1.4
Poller Type - Cactid v
Server Info - Linux 2.6.18-5-686
Web Server - Apache/2.2.3 (Debian) mod_python/3.2.10 Python/2.4.4 PHP/5.2.0-8+etch9 mod_ssl/2.2.3 OpenSSL/0.9.8c mod_perl/2.0.2 Perl/v5.8.8
PHP - 5.2.0-8+etch9
PHP Extensions - zip, xmlwriter, libxml, dom, xmlreader, xml, wddx, tokenizer, sysvshm, sysvsem, sysvmsg, session, SimpleXML, sockets, soap, SPL, shmop, standard, Reflection, posix, mime_magic, mbstring, json, iconv, hash, gettext, ftp, filter, exif, dba, date, ctype, calendar, bz2, bcmath, zlib, pcre, openssl, apache2handler, memcache, curl, gd, mcrypt, mhash, mysql, mysqli, PDO, pdo_mysql, snmp
MySQL - 5.0.32-Debian_7etch4
RRDTool - 1.2.15
SNMP - 5.2.3
Plugins
Device Monitoring (monitor - v0.8.1)
Global Plugin Settings (settings - v0.3)
Thresholds (thold - v0.3.
Host Info (hostinfo - v0.2)
Network Tools (tools - v0.2)

Any idea?

guanno · Post by **guanno** » Tue Oct 28, 2008 9:12 am

I foud a very simple reason for choppy graphs in that some of the devices we were monitoring had there SNMP settings changed. This caused a time out issue when querying them triggering timeouts on other device queries.Once the offending devices were fixed the choppiness went away.

MrJake · Post by **MrJake** » Fri Nov 07, 2008 3:34 pm

I redid all my snmp config in my 10 devices then resynced the devices in cacti and im still getting choppy graphs from all my URBs. only device giving me nice graphs are my 3750 and my 4507 (cisco switches). But all the rest of my graphs for ubr traffic, snr, fec whatever are choppy. this happened since I ran out of disk space on the partition which my rrd files are stored. before that day I never had choppy graphs. I even flushed all my graphs and data sources to star tfrom scratch. could it be the mysql db of cacti which is currupted or something?

i really love cacti but I would hate to have to dump it for another solution because until this choppy graphs shit this was working great

Post by **TheWitness** » Fri Nov 07, 2008 4:50 pm

So, what is in your log's? What poller are you running?

TheWitness

MrJake · Post by **MrJake** » Fri Nov 07, 2008 4:59 pm

my poller is spine here are my logs

11/07/2008 04:50:10 PM - SYSTEM STATS: Time:9.3993 Method:spine Processes:3 Threads:5 Hosts:11 HostsPerProcess:4 DataSources:1674 RRDsProcessed:1030
11/07/2008 04:45:11 PM - SYSTEM STATS: Time:9.2240 Method:spine Processes:3 Threads:5 Hosts:11 HostsPerProcess:4 DataSources:1674 RRDsProcessed:1030

if you read my first post i have the details there. but im thinking of scratching and restarting but the problem is the templates for the cisco ubr i dont think i have the patience to redo since it took for ever to make my host template top include all i needed

Post by **TheWitness** » Fri Nov 07, 2008 5:28 pm

If you dump your database and re-import it on the other system, then you are fine. You don't loose anything. You may have damaged something when your disk filled up. There is a repair script in the

Code: Select all

cli

directory though, you might try that, but if your RRDfiles are corrupt, well, then you have a different problem.

TheWitness

Post by **TheWitness** » Fri Nov 07, 2008 5:30 pm

Oh, one more thing, although if you are not getting warning's it may not help, reduce the max_oid's per get request to 1 for each problematic device and see what happens.

TheWitness

MrJake · Post by **MrJake** » Mon Nov 10, 2008 4:48 pm

I already set it to 1 hehe. Im gonna check out this repair script and also do a dump of the db. I will also flsuh every devices that have choppy graphs then de sure the rrd files are gone and start them over. Ill keep you posted sir

choppy graphs

choppy graphs

choppy graphs

Who is online