Think i screwed up a device template

[mrossTTX]

i had everything working well.. and i decided to get cute and downloaded a template to go after our HP blade servers.
dont ask me how.. but it seems that after i imported it, i lost the hard drive, memory and CPU graphing ability on windows hosts.

im running cacti 1.2.23, any way i can get the default template that was with the base install so i can remove the one i brought in for the blades and get back the default template? appreciate any help thanks.

btw, logged in users, uptime, NICs, all work.. just not the cpu, memory and disks.

[TheWitness]

Just re import the Windows device package. They are in the install/templates directory.

Console > Import/Export > Import Package

[mrossTTX]

thanks for that.. i re imported it.. but the disk graph is still not working, nor is the CPU.. it stopped working like 2 days ago.. when i edit the device, i can see that it can detect items for the disk\cpu.. i can create the graphs, but they're empty.. literally the disk says "0"..

[TheWitness]

Put the device into debug and look what coming in from the poller.

[mrossTTX]

here is what i got in the log right after putting it in debug mode

2023-01-27 07:00:49 - SPINE: Poller[Main Poller] PID[1348] PT[42950488912] Device[SERVERNAME] HT[1] DS[SERVERNAME - Traffic - 172.16.16.16 - ethernet_32773] Graphs[SERVERNAME - Traffic - Ethernet0 2] TT[15969.25] SNMP: v2: 172.16.16.16, dsname: traffic_out, oid: .1.3.6.1.2.1.2.2.1.16.9, value: 42022318
2023-01-27 07:00:49 - SPINE: Poller[Main Poller] PID[1348] PT[42950488912] Device[SERVERNAME] HT[1] DS[SERVERNAME - Used Space - C: Label: Serial Number 5a9e6322] Graphs[SERVERNAME - Used Space - C: Label: Serial Number 5a9e6322] TT[15954.65] SS[1] SERVER: C:\inetpub\wwwroot\cacti\scripts\ss_host_disk.php ss_host_disk "172.16.16.16" "529" "2:1615:1:o/GL2Wjb::::::" "get" "total" "1", output: 0
2023-01-27 07:00:33 - SPINE: Poller[Main Poller] PID[1348] PT[42950488912] Device[SERVERNAME] HT[1] DS[SERVERNAME - Used Space - C: Label: Serial Number 5a9e6322] Graphs[SERVERNAME - Used Space - C: Label: Serial Number 5a9e6322] TT[12047.10] SS[1] SERVER: C:\inetpub\wwwroot\cacti\scripts\ss_host_disk.php ss_host_disk "172.16.16.16" "529" "2:1615:1:o/GL2Wjb::::::" "get" "used" "1", output: 0
2023-01-27 07:00:21 - SPINE: Poller[Main Poller] PID[1348] PT[42950488912] Device[SERVERNAME] HT[1] DS[SERVERNAME - CPU Utilization - CPU1] Graphs[SERVERNAME - CPU Utilization - CPU1] TT[30.20] SS[0] SERVER: C:\inetpub\wwwroot\cacti\scripts\ss_host_cpu.php ss_host_cpu "172.16.16.16" "529" "2:1615:1:o/GL2Wjb::::::" "get" "usage" "1", output: 0
2023-01-27 07:00:21 - SPINE: Poller[Main Poller] PID[1348] PT[42950488912] Device[SERVERNAME] HT[1] DS[SERVERNAME - Traffic - 172.16.16.16 - ethernet_32773] Graphs[SERVERNAME - Traffic - Ethernet0 2] TT[2.49] SNMP: v2: 172.16.16.16, dsname: traffic_in, oid: .1.3.6.1.2.1.2.2.1.10.9, value: 56782452
2023-01-27 07:00:21 - SPINE: Poller[Main Poller] PID[1348] PT[42950488912] Device[SERVERNAME] HT[1] DS[SERVERNAME - Processes] Graphs[SERVERNAME - Processes] TT[6.53] SNMP: v2: 172.16.16.16, dsname: proc, oid: .1.3.6.1.2.1.25.1.6.0, value: 101
2023-01-27 07:00:08 - SPINE: Poller[Main Poller] PID[1348] PT[42950488912] Device[SERVERNAME] HT[1] DS[SERVERNAME - Logged in Users] Graphs[SERVERNAME - Logged in Users] TT[4.17] SNMP: v2: 172.16.16.16, dsname: users, oid: .1.3.6.1.2.1.25.1.5.0, value: 7
2023-01-27 07:00:07 - SPINE: Poller[Main Poller] PID[1348] PT[42950488912] DEBUG: Device[SERVERNAME] HT[1] RECACHE: Processing 2 items in the auto reindex cache for '172.16.16.16'
2023-01-27 07:00:07 - SPINE: Poller[Main Poller] PID[1348] PT[42950488912] Device[SERVERNAME] DEBUG: Attempting to ping 172.16.16.16, seq 4 (Retry 0 of 5)

[mrossTTX]

looked at the debug on the graph itself .. saw this.. its only the CPU and Disk graphs that I am having issues with at the moment.

[mrossTTX]

I put the device itself in debug mode (mangment\devices\devicename\click enable device debug)
then i waited for the log to populate (filtering by the device IP) and what I got in the log is what I posted: (what steps exactly, did I not listen to?)

2023-01-27 07:00:49 - SPINE: Poller[Main Poller] PID[1348] PT[42950488912] Device[SERVERNAME] HT[1] DS[SERVERNAME - Used Space - C: Label: Serial Number 5a9e6322] Graphs[SERVERNAME - Used Space - C: Label: Serial Number 5a9e6322] TT[15954.65] SS[1] SERVER: C:\inetpub\wwwroot\cacti\scripts\ss_host_disk.php ss_host_disk "172.16.16.16" "529" "2:1615:1:o/GL2Wjb::::::" "get" "total" "1", output: 0
2023-01-27 07:00:33 - SPINE: Poller[Main Poller] PID[1348] PT[42950488912] Device[SERVERNAME] HT[1] DS[SERVERNAME - Used Space - C: Label: Serial Number 5a9e6322] Graphs[SERVERNAME - Used Space - C: Label: Serial Number 5a9e6322] TT[12047.10] SS[1] SERVER: C:\inetpub\wwwroot\cacti\scripts\ss_host_disk.php ss_host_disk "172.16.16.16" "529" "2:1615:1:o/GL2Wjb::::::" "get" "used" "1", output: 0
2023-01-27 07:00:21 - SPINE: Poller[Main Poller] PID[1348] PT[42950488912] Device[SERVERNAME] HT[1] DS[SERVERNAME - CPU Utilization - CPU1] Graphs[SERVERNAME - CPU Utilization - CPU1] TT[30.20] SS[0] SERVER: C:\inetpub\wwwroot\cacti\scripts\ss_host_cpu.php ss_host_cpu "172.16.16.16" "529" "2:1615:1:o/GL2Wjb::::::" "get" "usage" "1", output: 0

[TheWitness]

Okay, my apologies, write it off to not enough coffee.

[mrossTTX]

hey its all good.. we have those days right? LOL!

to add further mystery.. the cacti server can see its own disk and CPU.. i cant see remote machine's disk and CPU.. but i can see remote machines uptime, processes, NIC information, etc.. just CPU and Disk is fubared..

[TheWitness]

Okay, there is one nasty bit about windows and SNMP, wien you say drop a CD in the CD drive, Microsoft switches the indexes. Their SNMP support has been broken since I started using cacti in 2002. Makes me quite angry.

So, perform a verbose query and check the indexes and make sure they did not change for the drives in question.

Also, I thought I updated those template scripts to pass in just the device id. But looking at my system, I guess not.

Anyway, likely the indexes changed. Maybe CPU too. Hopefully you understand this.

[TheWitness]

On the other machines, do a verbose query, and increase the SNMP timeout. Microsoft SNMP can be real slow especially with regard to disk stats.

[mrossTTX]

i MAY have found something (btw.. it wasnt just remote windows machines, also had the issue on ESXi)

i went to the HMIB setup in Settings and changed the storage and CPU from 4 hours to 5 minutes.
Mind, you , it was 4 hours for the past few days\weeks.. and it just stopped.. i am now seeing the disk on one remote machine. i will report back if i begin seeing the CPU

[TheWitness]

Lastly, if you use HMIB, you won't have these problems. It does async collection and reports on quite a bit more, but be careful with some of the frequencies as some run a long time, especially if you don't use snmpv2.

[mrossTTX]

Looks like CPU and DISK is now working since i changed the setting for HMIB.. odd tho as the cacti's graph for its own disk looks like "Classic" color scheme for graphs, but once I got HMIB set to 5min for disk and CPU, the graphs started working but the colors are different.. both servers appear to be using the same data query.. you didnt have enough coffee this morning.. i may below on sugar.. i'll keep digging.

graphChanges.JPG (57.2 KiB) Viewed 1334 times

[TheWitness]

I was going to say, "don't import ALL the templates from HMIB", the base ones in Cacti are better. So, you know the fix now right, import package and re-import only the Templates you want. If they come out janked, then remove orphan Graph Template Items during the package import.