Need help with abnormal graph

[ultmarksman]

My graph using custom template is becoming abnormal and I can't find the reason.
It looks like the time frame for calculation is sliding a little bit.
Please help.
Btw, sry for my bad english.

Cacti 0.8.8h
Spine 0.8.8h
poller/cron frequency 1 minute
rrdtool 1.3.8
Centos 6.8 64bit

/usr/bin/rrdtool graph - \
--imgformat=PNG \
--start='1488122239' \
--end='1488122934' \
--title='x' \
--rigid \
--base='1000' \
--height='120' \
--width='700' \
--alt-autoscale \
COMMENT:"From 2017/02/26 22\:17\:19 To 2017/02/26 22\:28\:54\c" \
COMMENT:" \n" \
--vertical-label='bits per second' \
--slope-mode \
--font TITLE:10: \
--font AXIS:8: \
--font LEGEND:8: \
--font UNIT:8: \
DEF:a='/var/www/html/cacti/rra/x_traffic_in_3772.rrd':'traffic_in':MAX \
DEF:b='/var/www/html/cacti/rra/x_traffic_in_3772.rrd':'traffic_out':MAX \
DEF:c='/var/www/html/cacti/rra/x_traffic_in_3772.rrd':'traffic_in':AVERAGE \
DEF:d='/var/www/html/cacti/rra/x_traffic_in_3772.rrd':'traffic_out':AVERAGE \
CDEF:cdefa='a,8,*' \
CDEF:cdefe='b,8,*' \
CDEF:cdefi='c,8,*' \
CDEF:cdefbe='d,8,*' \
LINE1:cdefa#FF0000FF:'' \
LINE1:cdefa#FF0000FF:'' \
LINE1:cdefa#FF0000FF:'Inbound Max\:' \
GPRINT:cdefa:MAX:' %8.2lf %s\n' \
LINE1:cdefe#FFC73BFF:'' \
LINE1:cdefe:'':STACK \
LINE1:cdefe#FFC73BFF:'Outbound Max\:' \
GPRINT:cdefe:MAX:'%8.2lf %s\n' \
AREA:cdefi#00FF00FF:'' \
AREA:cdefi:'':STACK \
AREA:cdefi#00FF00FF:'Inbound' \
GPRINT:cdefi:LAST:' Current\:%8.2lf %s' \
GPRINT:cdefi:AVERAGE:'Average\:%8.2lf %s' \
GPRINT:cdefi:MAX:'Maximum\:%8.2lf %s\n' \
LINE1:cdefbe#002A97FF:'' \
LINE1:cdefbe:'':STACK \
LINE1:cdefbe#002A97FF:'Outbound' \
GPRINT:cdefbe:LAST:' Current\:%8.2lf %s' \
GPRINT:cdefbe:AVERAGE:'Average\:%8.2lf %s' \
GPRINT:cdefbe:MAX:'Maximum\:%8.2lf %s\n'

[cigamit]

That certainly does appear a bit odd. Its not a counter overrun, since it is affecting both inbound and outbound and it should happen less frequently as the data amount gets smaller. I assume there isn't anything abnormal with your polling times (maybe check the timestamps for jumps around that period that it happens)? Is this a VM or Physical machine? Does the clock time appear normal or does it slowly get off?

[ultmarksman]

It is a physical machine. There is ntpd service running. The system time is fine as of now.

[ultmarksman]

below is data from rrd file using rrddump

<!-- 2017-02-26 22:20:00 ICT / 1488122400 --> <row><v> 1.5791285746e+08 </v><v> 1.0633301127e+09 </v></row>
<!-- 2017-02-26 22:21:00 ICT / 1488122460 --> <row><v> 1.7525291126e+08 </v><v> 1.0387343713e+09 </v></row>
<!-- 2017-02-26 22:22:00 ICT / 1488122520 --> <row><v> 1.6013074699e+08 </v><v> 1.0481739818e+09 </v></row>
<!-- 2017-02-26 22:23:00 ICT / 1488122580 --> <row><v> 2.4181889192e+08 </v><v> 1.6514525629e+09 </v></row>
<!-- 2017-02-26 22:24:00 ICT / 1488122640 --> <row><v> 6.4965584374e+07 </v><v> 4.5261781708e+08 </v></row>
<!-- 2017-02-26 22:25:00 ICT / 1488122700 --> <row><v> 1.4380629585e+08 </v><v> 1.0308955000e+09 </v></row>

and data from graph csv export

26-02-17 22:20 1263302859.70 8506640901.80
26-02-17 22:21 1402023290.00 8309874970.10
26-02-17 22:22 1281045976.00 8385391854.10
26-02-17 22:23 1934551135.40 13211620503.00
26-02-17 22:24 519724674.99 3620942536.60
26-02-17 22:25 1150450366.80 8247163999.70

[ultmarksman]

After I did a little searching. I think the problem may come from wrong RRA template settings. So I'm gonna test something out and I will tell the result later.

Btw, please feel free to comment if I'm doing something wrong. I'm new at cacti.