Patches item #694183, was opened at 2003-02-27 06:16
Message generated for change (Comment added) made by tbe
You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=301210&aid=694183&group_id=1210
Category: None
Group: None
>Status: Closed
>Resolution: Invalid
Priority: 5
Submitted By: Richard Scheffenegger (realarms)
Assigned to: Nobody/Anonymous (nobody)
Summary: Multiple patches and some minor bug fixes for 1.0.4-pre2
Initial Comment:
Generic CDEF datasource support fixed
-> some code in collector / compile tried to make
RRD-Datafile Sources from CDEFs
-> added the "PatchB" as found on the contributed
software site to grapher.cgi
Thanks to Kamil Vojtisek for his patch. See:
http://www.certaintysolutions.com/tech-
advice/cricket-contrib/files/PatchB.zip
-> Also very much of the same patch was reinvented
by Michael O'Reilly (Patch ID: 462714)
I cann't say now, were I took most of my code from,
just another example for
reinventing the wheel.
-> patch could be improved to use up less
RRDs::graph DEFs, but I figure if we do that
in perl, the gain in processing speed in RRDs::graph
would not be as significant
as the loss due to the more complex pre-processing
Added "Expires", "Refresh", "Last-Modified" Headers to
grapher.cgi and mini-graph.cgi, so that refreshing works.
=> will make shift an initialization for this value to conf-
cricket.pl,
since it's used on other places as well (like finding
out if a cached
graph can still be used).
-> just found that there was already a patch from Matt
Zimmerman (Patch ID: 480694)
to do almost the same (Refresh HTTP option is
missing).
Modified the use-gprint option
-> grapher.cgi also scans the local "graph" directive
for an entry
-> value not longer set to true / false, but instead it
gets scanned for
special characters:
"g" uses "grouping", ie the GPRINT commands
follow immediately to
the current graph legend and not using up
more space
"!" will put the last value GPRINT on the stack
"-" stands for the minimum value of the graph
"~" is for the average
"+" displays the maximum
-> Also the ordering of the last 4 of these options is
important, since they
also decide in what order the GPRINT values appear
below the graph.
Ie. "g-~+" will show a line like this in the legend part
of the graph
[_] legend Min: 0 Avg: 1 Max: 3
-> Added some code to fill up the "legend"s of all
datasources to be of equal length
so that mtargets with the gprint option look good
Fixed the formatting of the legend
-> code to put in \l options when necessary (end of
line)
-> removed some COMMENT\s statements
Added a new option "gprint-ds" to the graph directive
-> see my example for "pingperf.pl" on the contributed
page
-> makes the GPRINTs use their own datasource
instead of the
current datasource used to draw the graph. Added
this to
add flexibility with the CDEF datasources (ie. when
you stack
values as in my example, you end up drawing
calculated values,
but GPRINT should state the uncalculated values).
Fixed code for named events
-> Event will only show up in the graph and the legend
for that matter,
when it happened within the reach of the current
graph.
Fixed interaction of GPRINT / Events option
-> using both doesn't mess up the rrd-graph output
-> also fixed the "zero-rules" to go into the "@vrules"
array, this
also takes care of some formatting in the legend
Added special keyword "none" to summary-loc.
-> when using gprints / grouping, i figure an additional
summary can be omitted,
especially since the gprints from RRD are more
freely configurable.
Modified appearance of the Summary table
-> vertical alignement is now top instead of center
(handy for large mtarget's)
Minor fix for an uninitialized variable "$html" in
grapher.cgi/htmlCurrentPath
-> error would be logged if you enable the "show-path"
and were on the
cricket root directory. (main page)
Minor fix for an uninitialized variable "$rrd" in grapher.cgi
-> Happens when displaying mtargets; just added
sanity check for $rrd.
Modified the output for the warning after a RRD-Graph
error
-> the array will now be joined by "\n\t", so reading it
is much easier
----------------------------------------------------------------------
>Comment By: Terje Bless (tbe)
Date: 2005-02-27 17:19
Message:
Logged In: YES
user_id=8470
Patch missing and no followup from submitter. Closing as
invalid.
Please reopen, resubmit, or otherwise provide a heartbeat if
this
patch is still relevant.
----------------------------------------------------------------------
Comment By: Terje Bless (tbe)
Date: 2003-10-05 17:31
Message:
Logged In: YES
user_id=8470
This Patch Tracker Item has no actual patch file attached.
Also, this patch as described desperately needs to be split
up into manageable chunks before anyone in their right mind
will dare take a look at it. The plethora of stuff going on
(according to the description) will make developers
reluctant to use the scarce resource of their available time
on it.
Resubmit in more manageable chunks please, Richard? Add
separate patches (and as separate tracker items with
separate descriptions) so it's not so much to swallow at one go.
----------------------------------------------------------------------
You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=301210&aid=694183&group_id=1210
|