You are here

CANNOT CHOOSE HOST & SERVICE CHECK IN PERFORMANCE GRAPH APTER UPGRADE

7 posts / 0 new
Last post
HungNguyen
HungNguyen's picture
CANNOT CHOOSE HOST & SERVICE CHECK IN PERFORMANCE GRAPH APTER UPGRADE

I have upgraded to the lastest version of opsviews as below:

opsview-timeseries-rrd-1.0.0.170401507-1.ct7.x86_64

opsview-base-5.3.0.170401146-1.ct7.x86_64

opsview-timeseries-1.0.0.170261142-1.ct7.x86_64

opsview-timeseries-lib-1.0.0.170241408-1.ct7.x86_64

opsview-5.3.0.170791028-1.ct7.x86_64

mod_auth_tkt_opsview-2.1.4-1.el7.centos.x86_64

opsview-perl-5.3.0.170251214-1.ct7.x86_64

opsview-setup-1.0.0.162461538-1.ct7.x86_64

opsview-core-5.3.0.170791028-1.ct7.x86_64

opsview-web-5.3.0.170791028-1.ct7.x86_64

opsview-timeseries-enqueuer-1.0.0.170261221-1.ct7.x86_64

opsview-watchdog-1.0.0.163221006-1.ct7.x86_64

opsview-external-1.0.0.162792117-1.ct7.x86_64

But after that, I cannot choose host and service check in performance graph. The older graphs also doesn't render. It say: "... series that have not been found"

Please help me.

Thanks

Duncan Ferguson
dferguson's picture
This sounds like the

This sounds like the timeseries daemons are not running properly or the opsview-web process cannot talk to the daemon properly.

Does running 'opsview_watchdog summary' as the nagios user show all processes are 'Monitored' and 'Running'?

Do you have any proxies set up in your environment for the nagios user (check using 'env | grep -i proxy')?

If you do a reload, do you get any errors and do you see graphing icons in the UI?

  Duncs

HungNguyen
HungNguyen's picture
Hi Duncs,

Hi Duncs,

Thanks for your feedback, all processes are fine as below:

 

After reloading, I don't see any error, and I can see graphing icons:

Duncan Ferguson
dferguson's picture
Please download https://s3-eu

Please download https://s3-eu-west-1.amazonaws.com/opsview-patches/update-missing-timese..., switch to the opsview user (note, not nagios) and run

./update-missing-timeseries-uoms.pl /opt/opsview/timeseriesrrd/var/data

After this, perform a reload and check to see if the graphs are working.

  Duncs

HungNguyen
HungNguyen's picture
Hello

Hello

I have executed as bellow:

# sudo -iu opsview

$ wget https://s3-eu-west-1.amazonaws.com/opsview-patches/update-missing-timese...

$ chmod +x update-missing-timeseries-uoms.pl

$ ./update-missing-timeseries-uoms.pl /opt/opsview/timeseriesrrd/var/data

$ cat /opt/opsview/timeseriesrrd/var/data/+metadata.db 

Atableuomsuoms CREATE TABLE uoms (

            host VARCHAR(255) NOT NULL,

            service VARCHAR(255) NOT NULL,

            metric VARCHAR(255) NOT NULL,

            dstype VARCHAR(255) NOT NULL,

            uom VARCHAR(255) NOT NULL,

            PRIMARY KEY(host, service, metric)

       )'indexsqlite_autoindex_uoms_1uoms

Then, reload, but the graphs are still not show

I have tried to upgrate to 5.4 version, it excecute successfully but the graphs are still not show

Please advise me,

Thanks

HungNguyen
HungNguyen's picture
Update info:

Update info:

+ import_perfdatarrd is running:

# ps aux |grep import_perfdatarrd

nagios    7658  0.0  0.5 191332 19756 ?        Ss   Jul12   0:16 import_perfdatarrd

root     16077  0.0  0.0 112648   972 pts/1    S+   09:39   0:00 grep --color=auto import_perfdatarrd

+ There have some errors in opsviewd.log about import_perfdatarrd:

# grep import_perfdatarrd /var/log/opsview/opsviewd.log

[2017/07/12 11:39:53] [import_perfdatarrd] [Opsview::Utils::PerfdatarrdImporter::catch {...} :69] [ERROR] Processing file=1499834392.143236: Timeseries ERROR: {"errors":["Updater http://127.0.0.1:1620/ failed: Connection refused"]} at /opt/opsview/perl/lib/perl5/Opsview/Timeseries/Client.pm line 176.

[2017/07/12 14:37:22] [import_perfdatarrd] [main:::42] [INFO] Starting

[2017/07/12 14:37:22] [import_perfdatarrd] [main:::51] [FATAL] import_perfdatarrd already running as process 714

[2017/07/12 14:37:22] [import_perfdatarrd] [main::cleanup:139] [INFO] Stopping import_perfdatarrd - pid 

[2017/07/12 15:05:44] [import_perfdatarrd] [Opsview::Utils::PerfdatarrdImporter::catch {...} :69] [ERROR] Processing file=1499846742.048170: Timeseries ERROR: Failed to send HTTP request: Connection refused at /opt/opsview/perl/lib/perl5/Opsview/Timeseries/Client.pm line 163.

[2017/07/12 15:05:46] [import_perfdatarrd] [main::catch_signals:181] [WARN] Received kill signal - gracefully shutting down

[2017/07/12 15:05:46] [import_perfdatarrd] [main::cleanup:139] [INFO] Stopping import_perfdatarrd - pid 0

[2017/07/12 15:06:05] [import_perfdatarrd] [main:::42] [INFO] Starting

[2017/07/12 15:06:05] [import_perfdatarrd] [main:::70] [INFO] Daemonised

[2017/07/12 15:13:18] [import_perfdatarrd] [main::catch_signals:181] [WARN] Received kill signal - gracefully shutting down

[2017/07/12 15:13:18] [import_perfdatarrd] [main::cleanup:139] [INFO] Stopping import_perfdatarrd - pid 0

[2017/07/12 15:15:47] [import_perfdatarrd] [main:::42] [INFO] Starting

[2017/07/12 15:15:47] [import_perfdatarrd] [main:::70] [INFO] Daemonised

Duncan Ferguson
dferguson's picture
THis error is relevant:

This error is relevant:

Updater http://127.0.0.1:1620/ failed: Connection refused

On the server can you check this port is listening and can be communicated with (run these as the nagios and opsview users):

netstat -na | grep 1620
telnet localhost 1620
env | grep -i proxy

Also check your syslog/messages file for any errors from the timeseries daemons.

You can also restart all timeseries daemons by running as nagios:

for p in opsview-timeseriesinfluxdbupdates opsview-timeseriesinfluxdbqueries opsview-timeseriesenqueuer opsview-timeseries; do opsview_watchdog $p restart; done

 

  Duncs