Upgrade to 1.2.17 from 1.2.14 lost unix ping graph

[Rno]

Hello,
since I upgrade from 1.2.14 to 1.2.17 my unix ping host graph is not working anymore.

The round robin in the RRD debug show this:
0 AVERAGE 1500 1054 1 0.5 NaN 0
1 AVERAGE 1346 878 30 0.5 129.586 0
2 AVERAGE 1445 212 120 0.5 129.586 0
3 AVERAGE 4380 3751 480 0.5 129.586 0
4 MIN 1500 954 1 0.5 NaN 0
5 MIN 1346 970 30 0.5 7.13 0
6 MIN 1445 407 120 0.5 7.13 0
7 MIN 4380 1276 480 0.5 7.13 0
8 MAX 1500 1100 1 0.5 NaN 0
9 MAX 1346 157 30 0.5 7.35 0
10 MAX 1445 1233 120 0.5 7.35 0
11 MAX 4380 4252 480 0.5 7.35 0
12 LAST 1500 1173 1 0.5 NaN 0
13 LAST 1346 397 30 0.5 7.14 0
14 LAST 1445 22 120 0.5 7.14 0
15 LAST 4380 272 480 0.5 7.14 0

But the last update is when I did the upgrade:
last_update 1621498145 [20/05/2021 10:09:05]

[Rno]

And here is the debug I got:
2021/05/25 14:26:07 - SPINE: Poller[Main Poller] PID[1923] DEBUG: The Value of Active Threads is 10 for Device ID 8205
2021/05/25 14:26:07 - SPINE: Poller[Main Poller] PID[1923] DEBUG: Device[CISCO] HT[1] DEBUG: HOST COMPLETE: About to Exit Device Polling Thread Function
2021/05/25 14:26:07 - SPINE: Poller[Main Poller] PID[1923] Device[CISCO] HT[1] Total Time: 0.047 Seconds
2021/05/25 14:26:07 - SPINE: Poller[Main Poller] PID[1923] Device[CISCO] HT[1] Device has no information for recache.
2021/05/25 14:26:07 - SPINE: Poller[Main Poller] PID[1923] Device[CISCO] DEBUG: ICMP Device Alive, Try Count:1, Time:7.2801 ms
2021/05/25 14:26:07 - SPINE: Poller[Main Poller] PID[1923] Device[CISCO] DEBUG: Received EINTR
2021/05/25 14:26:07 - SPINE: Poller[Main Poller] PID[1923] Device[CISCO] DEBUG: Received EINTR
2021/05/25 14:26:07 - SPINE: Poller[Main Poller] PID[1923] Device[CISCO] DEBUG: Received EINTR
2021/05/25 14:26:07 - SPINE: Poller[Main Poller] PID[1923] Device[CISCO] DEBUG: Attempting to ping 2.20.215.7, seq 365 (Retry 0 of 3)
2021/05/25 14:26:07 - SPINE: Poller[Main Poller] PID[1923] Device[CISCO] DEBUG: Entering ICMP Ping
2021/05/25 14:26:07 - SPINE: Poller[Main Poller] PID[1923] DEBUG: In Poller, About to Start Polling of Device for Device ID 8205

[Rno]

No one is using the template 'Unix Ping host' ?

I can't make it work since upgrade to 1.2.17, and it was before!!

[netniV]

Is it the template, or is it spine?

[Rno]

Hello,
It's the template, spine is working fine with the admin right (SUID root)

[netniV]

Sorry, you are assuming that the template has been updated during an update? I don't believe that happens unless you tick the box requesting to reset the templates. If it was the generally graphing subsystem, more graphs would be affected.

I will see about adding a test device with this template but it is more likely that the changes to spine may have had a knock on effect for some reason.

Have you tried rolling spine back a version or two for testing purposes?

[Rno]

Bot I'm not assuming anything!
I just find that the template dosen't work anymore,that's all !
Is said it's the template since you ask if it was the template or Spine. Spine is working fine wiht my other 1500 devices.
But only this function/template is not working.

And No I didn't upgrade any template at all.

Ok I will try on monday to downgrade spine, but since spine work for other template I don't think it's the problem.

And by the wy I try to remove the device, and re-create it, but that dosent solve thei ssue

[netniV]

OK so there doesn't appear to be anything wrong with the code or template on my installation. Can you run the Data Source Debug against the datasource and see if anything comes back as wrong? Just to be clear, we are talking about the Unix Ping Latency graph right?

[netniV]

I'm seeing zero issues at this end. It would definitely seem to be environmental (firewall or permissions maybe?).

[Rno]

Well it's not since in CLI it's working.
But I have no clue if I have to change some right somwhere (like it' has to be done for spine with ping ICMP).

The ping to check if the device is up (Availability/Reachability Options) is working, but not for the graph template using the perl script (/scripts/ping.pl )

I try again to delete de graph, and re-create it, but when I'm on the create graph page, an click create, the button is grayed, and then back to normal, but nothing happen
And I got this error int the console log of my browser

Code: Select all

Error: Syntax error, unrecognized expression: #
jquery.js:1677:8
Sizzle</Sizzle.error
http://cacti.lausanne.ch/cacti/include/js/jquery.js:1677:8
Sizzle</Sizzle.tokenize
http://cacti.lausanne.ch/cacti/include/js/jquery.js:2377:4
Sizzle</Sizzle.select
http://cacti.lausanne.ch/cacti/include/js/jquery.js:2838:20
Sizzle
http://cacti.lausanne.ch/cacti/include/js/jquery.js:894:9
find
http://cacti.lausanne.ch/cacti/include/js/jquery.js:3095:4
jQuery.fn.init
http://cacti.lausanne.ch/cacti/include/js/jquery.js:3205:14
jQuery
http://cacti.lausanne.ch/cacti/include/js/jquery.js:157:10
<anonyme>
http://cacti.lausanne.ch/cacti/graphs_new.php:61:20
mightThrow
http://cacti.lausanne.ch/cacti/include/js/jquery.js:3762:21
resolve/</process<
http://cacti.lausanne.ch/cacti/include/js/jquery.js:3830:12

[netniV]

So I have seen that type of error before and that basically means some of the javascript code has tried to find an element, but failed to find the correct name. However, it doesn't check that the name is empty and then tries to find the element without supplying a name. All elements are named #myelementname or #this_is_my_id, so as you can imagine, looking for just # means ... well nothing and causes the error.

Definitely an issue, just not sure why that would have any bearings on the rest unless it's simply preventing the display just to the JS error. Can you attempt to locate that line of code for the graph_new.php entry ?

[Rno]

the line given in graph_new.php is:
var passwordMatchTooShort='Passphrase matches but too short';

So nothing to do with the error!


But How is it possible that grpah_new has to do with the fact that all graph created before the upgrade are still display NAN as value, instead of working like it was ?
I'm more concern by the fact the template is not working anymore, than the fact I can't creat graph of this type anymore, even if I think both are releated to the template, and not to the graph_new function.