Wrong values in latest data Zabbix 3.0.7

[sinai65]

Hi all,
I have a Zabbix server which contains some hosts (8 Switch devices and 40 Linux servers). Monitoring process is going to be good except in some cases I get the wrong values in Latest data.
I have created a script to get RAID status of my Linux servers and put it in zabbix_agentd.conf with the proper syntax. Until now everything is OK and when I check the status of a hard disk in RAID array with the use of zabbix_get command I can see the value "Online". The problem is that when I check the status of that Hard disk in the latest data, I see "Failed" value for it !!
The problem is that it is fluctuating, I mean when a check is occurred in Zabbix (1 per 5 minutes) it shows the value "Online" and after 2 or 3 next checks, it gets "Failed" and after some checks it shows "Online" again!!
I am using LLD for discovering Raid type and hard disks inside the arrays. This problem does not happen for all hosts only a few hosts are facing this issue.
Any help is appreciated.

[batchenr]

Hi all,
I have a Zabbix server which contains some hosts (8 Switch devices and 40 Linux servers). Monitoring process is going to be good except in some cases I get the wrong values in Latest data.
I have created a script to get RAID status of my Linux servers and put it in zabbix_agentd.conf with the proper syntax. Until now everything is OK and when I check the status of a hard disk in RAID array with the use of zabbix_get command I can see the value "Online". The problem is that when I check the status of that Hard disk in the latest data, I see "Failed" value for it !!
The problem is that it is fluctuating, I mean when a check is occurred in Zabbix (1 per 5 minutes) it shows the value "Online" and after 2 or 3 next checks, it gets "Failed" and after some checks it shows "Online" again!!
I am using LLD for discovering Raid type and hard disks inside the arrays. This problem does not happen for all hosts only a few hosts are facing this issue.
Any help is appreciated.

sometimes this error accoure when you have a device with too many unsupported items - if you see some in this specific device - disable them all.

what is the item interval ?

[sinai65]

There were some unsupported items already, but I removed them all and nothing changed!!
The item interval is set to 5 minutes. The wired thing is that when I check the status of a hard disk by zabbix_get command, it always return the right value (i.e. online) but zabbix front end shows the wrong value (i.e. faulty)!! I also changed the values to 1 and 0 instead of online and faulty, because I thought it is an issue with the string values but nothing changed.
Moreover, I checked the history_unit table for the stored values for this item and I understood the values are wrong in the database too!!
for example the latest data for one of my servers is as follows:

Name Last check Last Value
Hard No 2I:1:1 status 2017-02-22 08:31:14 Healthy (1)
Hard No 2I:1:2 status 2017-02-22 08:31:15 Faulty (0)
Hard No 2I:1:3 status 2017-02-22 08:31:17 Healthy (1) +1
Hard No 2I:1:4 status 2017-02-22 08:31:17 Healthy (1)

As you see it is shown that Faulty for "Hard No 2I:1:2 status" and, it was Faulty for "Hard No 2I:1:3 status" which is now Healthy!!! In fact, both of them should be Healthy.

[batchenr]

There were some unsupported items already, but I removed them all and nothing changed!!
The item interval is set to 5 minutes. The wired thing is that when I check the status of a hard disk by zabbix_get command, it always return the right value (i.e. online) but zabbix front end shows the wrong value (i.e. faulty)!! I also changed the values to 1 and 0 instead of online and faulty, because I thought it is an issue with the string values but nothing changed.
Moreover, I checked the history_unit table for the stored values for this item and I understood the values are wrong in the database too!!
for example the latest data for one of my servers is as follows:

Name Last check Last Value
Hard No 2I:1:1 status 2017-02-22 08:31:14 Healthy (1)
Hard No 2I:1:2 status 2017-02-22 08:31:15 Faulty (0)
Hard No 2I:1:3 status 2017-02-22 08:31:17 Healthy (1) +1
Hard No 2I:1:4 status 2017-02-22 08:31:17 Healthy (1)

As you see it is shown that Faulty for "Hard No 2I:1:2 status" and, it was Faulty for "Hard No 2I:1:3 status" which is now Healthy!!! In fact, both of them should be Healthy.

this values coming from the RAID script you created ?
if so can su to zabbix user at zabbix server - simply by
#su zabbix

and run the script with bash -x /scriptname.sh

and tell me if you saw some errors ?
zabbix server showing somthing ? maybe lost connection ?

[sinai65]

I am sending the output of running my script within the server and from zabbix server:

[root@irm ~]# zabbix_get -s serverip -k p410.hardcheck[2I:1:1]
1
[root@irm ~]# zabbix_get -s serverip -k p410.hardcheck[2I:1:2]
1
[root@irm ~]# zabbix_get -s serverip -k p410.hardcheck[2I:1:3]
1
[root@irm ~]# zabbix_get -s serverip -k p410.hardcheck[2I:1:4]
1

[root@myserver ~]# su zabbix --shell=/bin/bash

bash-4.1$ /usr/local/sbin/P410_Discover.sh Discover
{"data":[{"{#ARRAY}":"2I:1:1"},{"{#ARRAY}":"2I:1:2"},{"{#AR RAY}":"2I:1:3"},{"{#ARRAY}":"2I:1:4"}]}

bash-4.1$ /usr/local/sbin/P410_Discover.sh 2I:1:1
1
bash-4.1$ /usr/local/sbin/P410_Discover.sh 2I:1:2
1
bash-4.1$ /usr/local/sbin/P410_Discover.sh 2I:1:3
1
bash-4.1$ /usr/local/sbin/P410_Discover.sh 2I:1:4
1

As you see there is no error and there is nothing related to lost connection in zabbix server log files too!!