[alpha4] all disk relative items return 0

[PerroVerd]

Hi

I have set a host with a linux template, and all the disk items never change returning 0 values as seen on "Latest values"

Items returning 0
diskfree[/home]
diskfree_perc
diskused
diskused_perc
disktotal
inodefree
inodefree_perc
inodetotal

The items relative to disk diskread_bloks and diskwrite_blocks return values diferent than zero.

The hosts are a myriad of 2.4 an 2.6 kernels using ext3, reiserfs and in some cases lvm partitions, all of them running the zabbix_agentd.

Best regards

[Alexei]

Please, could you post content of /tmp/zabbix_agentd.tmp from one of monitored servers? Thanks.

[PerroVerd]

[alpha4] all disk relative items return 0

Of course, here it goes
The other hosts doesn't have any disk info in this file

osiris:~# cat /tmp/zabbix_agentd.tmp
netloadout1[lo] 106.133333
netloadout5[lo] 116.183333
netloadout15[lo] 115.994444
netloadin1[lo] 106.133333
netloadin5[lo] 116.183333
netloadin15[lo] 115.994444
netloadout1[eth0] 1924.533333
netloadout5[eth0] 2982.480000
netloadout15[eth0] 2431.626667
netloadin1[eth0] 8824.533333
netloadin5[eth0] 12323.733333
netloadin15[eth0] 10185.777778
netloadout1[eth1] 7735.466667
netloadout5[eth1] 10568.426667
netloadout15[eth1] 8573.902222
netloadin1[eth1] 1103.733333
netloadin5[eth1] 1887.520000
netloadin15[eth1] 1503.466667
netloadout1[teql0] 0
netloadout5[teql0] 0
netloadout15[teql0] 0
netloadin1[teql0] 0
netloadin5[teql0] 0
netloadin15[teql0] 0
cpu[idle1] 95.502508
cpu[idle5] 94.633167
cpu[idle15] 95.016321
cpu[user1] 2.629303
cpu[user5] 3.185566
cpu[user15] 2.921427
cpu[nice1] 0.294067
cpu[nice5] 0.306576
cpu[nice15] 0.311261
cpu[system1] 1.574122
cpu[system5] 1.874692
cpu[system15] 1.750991

Thanks in advance

[Alexei]

Hmm... I do not see any disk related metrics in the file. In this case, all the items must become UNSUPPORTED. I'll try to reproduce the problem tonight on my test system.

[Kayou]

Hello all,

Well i seem to meet the same problem, no information returning about disk space usage, though inode table information are fine.

Here zabbix_agentd.tmp on my server is returning the same info as PerroVerd.

I m running zabbix currently on suse 9.2.

Kayou

[den_RDC]

I also see 0 for disk (and inode related items (free total & used), but my zabbix_agentd log does show it returns the correct values...

031276:20050126:083800 Before read()
031276:20050126:083800 After read() 2 [12]
031276:20050126:083800 Got line:diskfree[/]
031276:20050126:083800 Sending back:19980560.000000

Inodes seem to return 0 though, could it have something to do with reiserfs?

031275:20050126:083800 Before read()
031275:20050126:083800 After read() 2 [13]
031275:20050126:083800 Got line:inodefree[/]
031275:20050126:083800 Sending back:0.000000

Using 1.1A4 on k2.6.8-x86

[Alexei]

I also see 0 for disk (and inode related items (free total & used), but my zabbix_agentd log does show it returns the correct values...

031276:20050126:083800 Before read()
031276:20050126:083800 After read() 2 [12]
031276:20050126:083800 Got line:diskfree[/]
031276:20050126:083800 Sending back:19980560.000000

What value of diskfree[/] do you see in ZABBIX frontend? 19980560 or 0?

[den_RDC]

What value of diskfree[/] do you see in ZABBIX frontend? 19980560 or 0?

0
And i check the item manually in the DB, and it's also 0 there... so my (very humble) opinion would that there is a bug somewhere in zabbix_server.

[shadowcaster]

freebsd 4.10/zabbix 1.1alpha5 server/1.1alpha4 client (alpha5 distro)
LATEST DATA
Free disk space on / 07 Feb 17:54:30 0 B - Graph
Free disk space on /home 07 Feb 17:54:30 0 B - Graph
Free disk space on /tmp 07 Feb 17:54:30 0 B - Graph
Free disk space on /usr 07 Feb 17:54:30 0 B - Graph
Free disk space on /var 07 Feb 17:54:30 0 B - Graph
Free number of inodes on / 07 Feb 17:54:01 128002 - Graph
Free number of inodes on /home 07 Feb 17:54:01 4898557 - Graph
Free number of inodes on /tmp 07 Feb 17:54:01 128002 - Graph
Free number of inodes on /usr 07 Feb 17:54:01 69081 +1 Graph
Total disk space on / 07 Feb 17:18:47 0 B - Graph
Total disk space on /home 07 Feb 17:18:46 0 B - Graph
Total disk space on /tmp 07 Feb 17:18:47 0 B - Graph
Total disk space on /usr 07 Feb 17:18:47 0 B - Graph
Total number of inodes on / 07 Feb 17:18:46 129790 - Graph
Total number of inodes on /home 07 Feb 17:18:47 5174398 - Graph
Total number of inodes on /tmp 07 Feb 17:18:47 129790 - Graph
Total number of inodes on /usr 07 Feb 17:18:46 232318 - Graph
Used disk space on / 07 Feb 17:54:30 0 B - Graph
Used disk space on /home 07 Feb 17:54:30 0 B - Graph
Used disk space on /tmp 07 Feb 17:54:30 0 B - Graph
Used disk space on /usr 07 Feb 17:54:30 0 B - Graph
Used disk space on /var 07 Feb 17:54:30 0 B - Graph

[Alexei]

I still have no news. Please, expect more information later today.

[Alexei]

I think I found what happens. SuSE 9.2 has ReiserFS installed by default. ReiserFS does not provide kernel with inode-related information. It just silently returns 0. Very weird!

As for disktotal[], diskused[] - all these work for me.

Please, confirm that:

1. Problem with inode statistics is with ReiserFS only
2. disktotal[], diskfree[] work as expected

[Saltiel]

same problem on debian

Hello,

I get the same problem on a debian woody. I get a zero value for monitored disks.

I tried the zabbix 1.0 2 days ago, and there was no problems.

The agent got the value:

010750:20050208:134435 Got line:diskfree[/usr]
010750:20050208:134435 Sending back:343868.000000

I'm sure the server is doing request to get back these values. But there is nothing in the db. I opened the sock and the port for mysql, everything seems to be correct in the config files, and there is nothing in the logs, even in debugging mode.

on the server:

014233:20050208:134430 get_value_agent: host[as] ip[192.168.7.2] key [diskfree[/]]
014233:20050208:134430 Sending [diskfree[/]
014233:20050208:134430 Got string:[14] [2209652.000000]
014233:20050208:134430 RESULT_STR [2209652.00000]
014233:20050208:134430 GOT VALUE [2209652.00000]
014233:20050208:134430 In process_new_value()
014233:20050208:134430 In add_history()
014233:20050208:134430 Executing query:insert into history (clock,itemid,value) values (1107866670,17210,0.000000)
014233:20050208:134430 In add_trend()
014233:20050208:134430 SQL [select num,value_min,value_avg,value_max from trends where itemid=17210 and clock=1107864000]
014233:20050208:134430 Executing query:select num,value_min,value_avg,value_max from trends where itemid=17210 and clock=1107864000
014233:20050208:134430 In DBnum_rows
014233:20050208:134430 Result of DBnum_rows [1]
014233:20050208:134430 Executing query:update trends set num=83, value_min=0.000000, value_avg=0.000000, value_max=0.000000 where itemid=17210 and clock=1107864000
014233:20050208:134430 Executing query:update items set nextcheck=1107866700,lastclock=1107866670 where itemid=17210

Any help would be greatly appreciated

[Alexei]

Please, check if you're using custom multiplier for the diskfree[/usr] and the multiplier is not equal to '0' (if you're).

Regards,
Alexei

[den_RDC]

Still the same problems with alpha 5. I knew inode statistics were not reported by ReiserFS (didn't i mention that in a previous post ?) but the diskfree/total/usage should work. I double checked there were no multipliers in my configuration at all.

[Saltiel]

Hey!

it works better

Thnx for the quick help