Cheshire Cat Computing

Basically, the bad example is missing out the lines entirely, the good example is not. However since you have 'active' set, in theory the first definition is more correct since the lines should be suppressed. The second version merely hides the labels but still outputs the text (I assume the label hiding isdue to your definitions).

RRDTool gets very sulky if there are no RRD files being graphed, eg in the case where lines are suppressed - this is why the first example fails. It should default to adding at least one hidden line if possible... I need to run some tests on this but at least I know where to look now. I'll release a 2.19b1 when I've verified it fixed

_________________
Steve Shipway
UNIX Systems, ITSS, University of Auckland, NZ
Woe unto them that rise up early in the morning... -- Isaiah 5:11

Is there a way to work around this based on the included file above (ie: cp-sw01.txt) ?
There are some features in 2.19 I'd really like to test out .
Tia , JimL

ps: Should have added this to the above request/comment , if at any time you need someone to test a possible fix Just shoot me a tarball or where to get it . @ babydr -at- baby-dragons -dot- com .

I'm at the LISA08 conference this week, so its a bit hard to work on this. I'll see what I can do but it probably needs to wait 10 days... I will let you know.

_________________
Steve Shipway
UNIX Systems, ITSS, University of Auckland, NZ
Woe unto them that rise up early in the morning... -- Isaiah 5:11

Just a reminder , Know you are busy . reply when time permits. Tnx , JimL

Tracked down the problem, and duplicated it here.

A minor 'optimisation' in 2.19 made sub-interfaces get skipped if they were 'inactive' and the isactive flag was set. Previously, the DEF options were all calculated anyway and the skip was between the DEF and the LINE options, so it worked but was lots more work for RRDTool.

I'm coding in a special bit to identify this case and greyout the graph instead.

[update]
Fix done and verified. If no sub-targets are active, you now get a greyed-out graph instead from 0-1 on the y-axis. I'll test a few more things before releasing a beta... you can have a copy now to help test if you'd like?

_________________
Steve Shipway
UNIX Systems, ITSS, University of Auckland, NZ
Woe unto them that rise up early in the morning... -- Isaiah 5:11

Hooray , router2.cgi doesn't spew all over the graphs now .

But , There is always one .

These greyed out graphs -Hopefully- we can get back our active/inactive status check so these greyed out graphs won't be shown ?

tia , JimL

2.20b1 has a fix in it for userdefined graphs where 'active' mode is set but no subtargets are active. It gives a completely greyed-out graph in these cases with no legend.

_________________
Steve Shipway
UNIX Systems, ITSS, University of Auckland, NZ
Woe unto them that rise up early in the morning... -- Isaiah 5:11

Hello Steve , Yes I am using 2.20b1 right now .
My problem is I have very large switches & my monitoring team is having difficulties going thru all those ports that are greyed out looking for the one they need .
Is there a method available to get back our "If there is no data for this target don't display it" , As we had before this change ? I'm game for most any work around .
Tia , JimL

Sorry, still trying to understand exactly what you're after.

Do you mean -
1) hide the Target from the menu if it is inactive? - This shouldnt normally happen; you can get this effect using (3) below though
2) hide the target from the summary graph if it is inactive? (grey out the summary if no active targets) - 2.19+ does this if 'active' option is set, apart from the bug if none are active which is fixed in 2.20b1
3) hide the target from the summary page if it is inactive? (2.19+ should do this if 'active' option set)

_________________
Steve Shipway
UNIX Systems, ITSS, University of Auckland, NZ
Woe unto them that rise up early in the morning... -- Isaiah 5:11