[SCM] BOINC packaging branch, master, updated. debian/7.0.33+dfsg-1-90-g33b5da5
Gianfranco Costamagna
costamagnagianfranco at yahoo.it
Mon Nov 5 12:17:12 UTC 2012
The following commit has been merged in the master branch:
commit 33b5da5ffda2a7fe3740aa05acff87dca2ae6d93
Author: Gianfranco Costamagna <costamagnagianfranco at yahoo.it>
Date: Mon Nov 5 13:16:50 2012 +0100
Another patch to fix a ftbfs
diff --git a/debian/changelog b/debian/changelog
index 3406e77..515c57a 100644
--- a/debian/changelog
+++ b/debian/changelog
@@ -3,8 +3,9 @@ boinc (7.0.38+dfsg-1) UNRELEASED; urgency=low
* New upstream version
* Revised some patches (some of them should will be removed in a future commit)
* Added add-missing-test-files.patch to fix a FTBFS
+ * Added add-missing-stripchart-samples.patch to fix another FTBFS
- -- Gianfranco Costamagna <costamagnagianfranco at yahoo.it> Mon, 05 Nov 2012 10:54:32 +0100
+ -- Gianfranco Costamagna <costamagnagianfranco at yahoo.it> Mon, 05 Nov 2012 13:16:32 +0100
boinc (7.0.36+dfsg-3) UNRELEASED; urgency=low
diff --git a/debian/patches/add-missing-stripchart-samples.patch b/debian/patches/add-missing-stripchart-samples.patch
new file mode 100644
index 0000000..90f53b6
--- /dev/null
+++ b/debian/patches/add-missing-stripchart-samples.patch
@@ -0,0 +1,473 @@
+Cherry picking commit 4904c8cc in order to add missing stripchart files (boinc ftbfs without them)
+--- /dev/null
++++ boinc-7.0.38+dfsg/stripchart/README
+@@ -0,0 +1,333 @@
++Stripchart version 2.0
++----------------------
++Author: Matt Lebofsky
++ BOINC/SETI at home - University of California, Berkeley
++ mattl at ssl.berkeley.edu
++
++Date of recent version: November 4, 2002
++
++Requirements:
++ * a gnuplot with the ability to generate gifs
++ * perl
++ * apache or other cgi-enabled web browser
++
++Send all thoughts and queries to: mattl at ssl.berkeley.edu
++
++This software is free to edit, distribute and use by anybody, as long as
++I get credit for it in some form or another. Thanks.
++----------------------
++
++Contents:
++
++I. Some questions and answers
++II. So how does it work?
++III. Known bugs, things to do, etc.
++
++----------------------
++I. Some questions and answers
++
++Q: What is stripchart?
++
++A: Well, it's actually two relatively small perl programs:
++
++ 1. stripchart
++
++ stripchart reads in time-based user data and, depending on a flurry of
++ command line options, generates a web-friendly .gif plotting the data.
++ The user can supply the time range, the y axis range, even the color
++ scheme, and more.
++
++ 2. stripchart.cgi
++
++ stripchart.cgi is a web-based GUI interface that allows users to easily
++ select multiple data sources and various parameters to plot, allowing
++ fast comparisons without having to deal with a command line interface.
++
++Q: Why do you bother writing this program?
++
++A: Working as a systems administrator (amongst other things) for SETI at home,
++ we kept finding ourselves in dire problem-solving situations, i.e. Why
++ did the database stop working? Why is load on our web server so high?
++
++ So we started collecting data in flat files, keeping track of server
++ loads, database checkpoint times, even CPU temperatures. When these files
++ grew too large and unwieldy, I found myself writing (and rewriting) simple
++ scripts to generate plots on this data. Sick of constant revision whenever
++ a new problem arose, I wrote stripchart version 1.0.
++
++ Its usefulness became immediately apparent when I added on stripchart.cgi.
++ I couldn't bear to teach everybody the many command line options to
++ stripchart, so I wrote this CGI to do all the dirty work. Suddenly we were
++ able to line up several plots, look for causes and effects, or just enjoy
++ watching the counts in our database tables grow to impossibly high numbers.
++
++ The SETI at home network has proven to be a delicate system, and keeping track
++ of all the data server, user, and web statistics has proven to be quite a
++ life saver. So when BOINC came around we felt that any project aiming to
++ embark on a similar project may need this tool. So I rewrote stripchart to
++ be a bit more friendly and general.
++
++Q: Why don't you make .pngs or .jpgs instead of .gifs? The latest gnuplot
++ doesn't support .gifs.
++
++A: Basically gnuplot support for other graphic file formats isn't as good. For
++ example, you cannot control exact window size, font size, and colors unless
++ you make .gifs. I'm not exactly sure why this is the case, but there you have it.
++ Anywho, you can find older gnuplot distributions out there - you'll need to
++ get the gd libs first, by the way.
++
++----------------------
++II. So how does it work?
++
++You can use stripchart as a stand alone command-line program to produce plots
++whenever you like, but we highly recommend using it in conjunction with the
++stripchart.cgi for ease of use. But here's how to do it both ways.
++
++stripchart (stand alone)
++
++Before anything, look at the section GLOBAL/DEFAULT VARS in the program
++stripchart and see if you need to edit anything (usually pathnames to
++executables and such).
++
++Let's just start with the usage (obtained by typing "stripchart -h"):
++
++stripchart: creates stripchart .gif graphic based on data in flat files
++options:
++ -i: input FILE - name of input data file (mandatory)
++ -o: output FILE - name of output .gif file (default: STDOUT)
++ -O: output FILE - name of output .gif file and dump to STDOUT as well
++ -f: from TIME - stripchart with data starting at TIME
++ (default: 24 hours ago)
++ -t: to TIME - stripchart with data ending at TIME (default: now)
++ -r: range RANGE - stripchart data centered around "from" time the size
++ of RANGE (overrides -t)
++ -l: last LINES - stripchart last number of LINES in data file
++ (overrides -f and -t and -r)
++ -T: title TITLE - title to put on graphic (default: FILE RANGE)
++ -x: column X - time or "x" column (default: 2)
++ -y: column Y - value or "y" column (default: 3)
++ -Y: column Y' - overplot second "y" column (default: none)
++ -b: baseline VALUE - overplot baseline of arbitrary value VALUE
++ -B: baseline-avg - overrides -b, it plots baseline of computed average
++ -d: dump low VALUE - ignore data less than VALUE
++ -D: dump high VALUE - ignore data higher than VALUE
++ -v: verbose - puts verbose runtime output to STDERR
++ -L: log - makes y axis log scale
++ -c: colors "COLORS" - set gnuplot colors for graph/axis/fonts/data (default:
++ "xffffff x000000 xc0c0c0 x00a000 x0000a0 x2020c0"
++ in order: bground, axis/fonts, grids, pointcolor1,2,3)
++ -C: cgi - output CGI header to STDOUT if being called as CGI
++ -s: stats - turn extra plot stats on (current, avg, min, max)
++ -j: julian times - time columns is in local julian date (legacy stuff)
++
++notes:
++ * TIME either unix date, julian date, or civil date in the form:
++ YYYY:MM:DD:HH:MM (year, month, day, hour, minute)
++ If you enter something with colons, it assumes it is civil date
++ If you have a decimal point, it assumes it is julian date
++ If it is an integer, it assumes it is unix date (epoch seconds)
++ If it is a negative number, it is in decimal days from current time
++ (i.e. -2.5 = two and a half days ago)
++ * All times on command line are assumed to be "local" times
++ * All times in the data file must be in unix date (epoch seconds)
++ * RANGE is given in decimal days (i.e. 1.25 = 1 day, 6 hours)
++ * if LINES == 0, (i.e. -l 0) then the whole data file is read in
++ * columns (given with -x, -y, -Y flags) start at 1
++ * titles given with -T can contain the following key words which will
++ be converted:
++ FILE - basename of input file
++ RANGE - pretty civil date range (in local time zone)
++ the default title is: FILE RANGE
++
++...okay that's a lot to ingest, but it's really simple. Let's take a look at an
++example (you'll find in the samples directory two files get_load and crontab).
++
++You have a machine that you want to monitor it's load. Here's a script that
++will output a single line containing two fields for time and the third with the
++actual data. For example:
++
++2002:11:05:12:51 1036529480 0.25
++
++The first field is time in an arbitrary human readable format
++(year:month:day:hour:minute), the second in epoch seconds (standard
++unix time format - the number of seconds since 00:00 1/1/1970 GMT),
++and the third is the load at this time.
++
++And we'll start collecting data every five minutes on this particular machine
++by add such a line to the crontab:
++
++0,5,10,15,20,25,30,35,40,45,50,55 * * * * /usr/local/stripchart/samples/get_load >> /disks/matt/data/machine_load
++
++So the file "machine_load" will quickly fill with lines such as the above.
++Now you may ask yourself - why two columns representing time in two different
++formats? Well sometime you just want to look at the data file itself, in which
++case the human-readable first column is quite handy to have around, but when
++making linear time plots, having time in epoch seconds is much faster to
++manipulate. So generally, we like to have at least the two time fields first,
++and the actual data in the third column. That's what stripchart expects by
++default.
++
++Note: stripchart will understand time in both epoch seconds and julian date.
++If the second time field is in julian date, you should supply the command line
++flag "-j" to warn stripchart so it knows how to handle it.
++
++Okay. So you have this data file now. A very common thing to plot would be the
++data over the past 24 hours. Turns out that's the default! If you type on the
++command line:
++
++stripchart -i machine_load -o machine_load.gif
++
++you will quickly get a new file "machine_load.gif" with all the goods.
++
++Note: you always have to supply an input file via -i. If you don't supply
++an output file via "-o" it .gif gets dumped to stdout. If you supply an
++output file via "-O" the output is stored in both the file and to stdout.
++
++Now let's play with the time ranges. You can supply times in a variety of
++formats on the command line:
++
++ "civil date" i.e. 2002:11:05:12:51 (YYYY:MM:DD:hh:mm)
++ "epoch seconds" i.e. 1036529480
++ "julian date" i.e. 2452583.52345
++
++You can supply a date range using the -f and -t flags (from and to):
++
++stripchart -i machine_load -f 2002:11:01:00:00 -t 2002:11:04:00:00
++
++Usually the "to" time is right now, so you can quickly tell stripchart
++to plot starting at some arbitrary time "ago." This is done also via the
++"-f" flag - if it's negative it will assume you mean that many decimal
++days from now as a starting point. So "-f -3.5" will plot from 3 and a
++half days ago until now.
++
++You can also supply a "range" centered around the from time. For example,
++to plot the 24 hours centered around 2002:11:01:13:40:
++
++stripchart -i machine_load -f 2002:11:01:13:40 -r 1
++
++On some rare occasions you might want to plot the last number of lines
++in a file, regardless of what time they were. If you supply the number
++of lines via the "-l" flag, it overrides any time ranges you may have
++supplied.
++
++Moving on to some other useful flags in no particular order:
++
++To change the default title (which is the basename of the file and
++the time range being plotted), you can do so via the "-T" command.
++Make sure to put the title in quotes. Within the title string the
++all-uppercase string "FILE" will be replaced with the file basename,
++and the string "RANGE" will be replaced by the time range. So in
++essence, the default title string is "FILE RANGE".
++
++If you have data files in different formats, you can specify the data
++columns using the "-x" and "-y" flags. By default -x is 2 and -y is 3.
++Sometimes we have datafiles with many columns so we actively have to tell
++stripchart which is the correct data column.
++
++However, you might want to overplot one column on top of another. If your
++data file has a second data column, you can specify what that is via the
++-Y flag, and this data will be overplotted onto the data from the first
++data column.
++
++Sometime you want to plot a horizontal rule or a "baseline". You can
++turn this feature on by specifying the value with the "-b" flag. If you
++use the "-B" flag (without any values) it automatically computes the
++average over the time range and plots that as the baseline. Simple!
++
++If you want to excise certain y values, you can do so with the dump
++flags, i.e. "-d" and "-D". In particular, any values lower than the one
++supplied with "-d" will be dumped, and any values higher supplied by
++"-D" will be dumped.
++
++To log the y axis, use the "-L" flag. Quite straightforward.
++
++A very useful flag is "-s" which outputs a line of stats underneath
++the plot title. It shows the current value, and the minimum, maximum
++and average values during the plot range.
++
++For verbose output to stderr, use the "-v" flag. It may not make much
++sense, but it's useful for debugging.
++
++Using the "-C" flag causes stripchart to spit out the "Content-type"
++lines necessary for incorporating stripchart plots into CGIs. This
++doesn't work so well now, but there it is.
++
++Okay. That's enough about the flags, and hopefully enough to get you
++playing around with stripchart and plotting some stuff. Now onto:
++
++stripchart.cgi
++
++First and foremost, you need to do the following before running the
++CGI version of stripchart:
++
++1. Put stripchart.cgi in a cgi-enabled web-accessible directory
++2. Make a "lib" directory somewhere that the web server can read/write to
++3. Edit stripchart.cgi GLOBAL/DEFAULT VARS to point to proper paths, including
++ the files "querylist" and "datafiles" in the aforementioned "lib" directory.
++4. Edit the "lib/datafiles" file to contain entries for all your data files.
++ You can find an example datafiles in the samples directory. Follow the
++ instructions in the comment lines, adding your entries below the header.
++
++That should be it, I think. Now go to the URL wherever your stripchart.cgi
++is sitting. If all is well..
++
++You will be immediately presented with a web form. Ignore the "select query"
++pulldown menu for now. Underneath that you will see a line:
++
++Number of stripcharts: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
++
++By default stripchart.cgi presents you with the ability to plot 4 simultaneous
++stripcharts, but you can select any number 1-20 by clicking on those numbers.
++The less plots, the faster a web page gets generated.
++
++For each plot, you get a pull down menu which should contain all the entries
++you already put in "datafiles". Here you are selecting your data source.
++
++Then you can select the time of time range: last x hours, last x days, or
++an arbitrary date range. By default the last x hours radio button is selected -
++to pick another type of time range make sure you select the radio button
++before it. Then enter the range via the pull down menus.
++
++Then you get a simple list of checkbox/input options. You can check to log
++the y axis, baseline the average, baseline an arbitrary value (which you
++enter in the window, enter a y minimum, or enter a maximum.
++
++When everything is selected, click on the "click here" button to plot.
++Depending on the speed of your machine, you should soon be presented with
++all the plots your desired, and the form underneath the plots which can
++edit to your heart's content. If you want to reset the form values, click
++on the "reset form" link.
++
++Note the "save images in /tmp" checkbox. If that is checked and you plot
++the stripcharts, numbered .gif files will be placed in /tmp on the web
++server machine so you can copy them elsewhere (files will be named:
++stripchart_plot_1.gif, etc.).
++
++On the topmost "click here" button you will note an "enter name to save
++query" balloon. If you enter a name here (any old string) this exact query
++will be saved into the "querylist" file which will then later appear in the
++pulldown menu at the top. That way if you have a favorite set of diagnostic
++plots which you check every morning, you don't have to enter the entire form
++every time.
++
++If you want to delete a query, enter the name in that same field but click
++the "delete" checkbox next to it. Next time you "click here" the query will
++be deleted.
++
++----------------------
++III. Known bugs, things to do, etc.
++
++* stripchart -C flag is kind of pointless and doesn't work in practice.
++* plots on data collected over small time ranges (points every few seconds, for
++ example) hasn't been tested.
++* plots that don't work via stripchart.cgi either show ugly broken image icons
++ or nothing at all - either way it's ungraceful.
++* pulldown menus and various plots sometimes need to be refreshed via a hard
++ refresh (i.e. shift-refresh).
++* this readme kinda stinks.
++* and many many other issues I'm failing to detail now!
++
++If you have any problems using the product, feel free to e-mail me at:
++
++ mattl at ssl.berkeley.edu
++
+--- /dev/null
++++ boinc-7.0.38+dfsg/stripchart/samples/get_load
+@@ -0,0 +1,11 @@
++#! /bin/csh
++
++set UPTIME = `/usr/bin/uptime | awk '{print $10}' | sed s/,//`
++if ($UPTIME == "average:") then
++ set UPTIME = `/usr/bin/uptime | awk '{print $11}' | sed s/,//`
++endif
++
++set CIVDATE = `/bin/date "+%Y:%m:%d:%H:%M"`
++set UNIXDATE = `/usr/local/bin/perl -e 'print time()'`
++
++echo $CIVDATE $UNIXDATE $UPTIME
+--- /dev/null
++++ boinc-7.0.38+dfsg/stripchart/samples/db_query
+@@ -0,0 +1,15 @@
++#! /bin/csh
++
++
++
++set CIVDATE = `date "+%Y:%m:%d:%H:%M"`
++
++set UNIXDATE = `perl -e 'print time()'`
++
++set DBNAME = `./parse_config db_name`
++
++set COUNT = `echo "use $DBNAME;$1" | mysql | tail -1`
++
++
++
++echo $CIVDATE $UNIXDATE $COUNT
+--- /dev/null
++++ boinc-7.0.38+dfsg/stripchart/samples/looper
+@@ -0,0 +1,6 @@
++#! /bin/csh
++
++while (1)
++ $1
++ sleep $2;
++end
+--- /dev/null
++++ boinc-7.0.38+dfsg/stripchart/samples/crontab
+@@ -0,0 +1 @@
++0,5,10,15,20,25,30,35,40,45,50,55 * * * * /usr/local/stripchart/samples/get_load >> /disks/matt/data/machine_load
+--- /dev/null
++++ boinc-7.0.38+dfsg/stripchart/samples/db_count
+@@ -0,0 +1,9 @@
++#! /bin/csh
++
++set CIVDATE = `date "+%Y:%m:%d:%H:%M"`
++set UNIXDATE = `perl -e 'print time()'`
++set DBNAME = `./parse_config db_name`
++
++set COUNT = `echo "use $DBNAME; select count(*) from $1" | mysql | tail -1`
++
++echo $CIVDATE $UNIXDATE $COUNT
+--- /dev/null
++++ boinc-7.0.38+dfsg/stripchart/samples/parse_config
+@@ -0,0 +1,40 @@
++#!/usr/bin/env python
++
++# -*- mode: python; python-indent: 4; -*-
++
++
++
++## $Id: parse_config 3581 2004-06-15 19:41:26Z rwalton $
++
++
++
++'''
++
++
++
++A program to pass config file settings to calling programs via stdout
++
++
++
++'''
++
++
++
++import boinc_path_config
++from Boinc import configxml
++
++import sys, os
++
++
++
++program_path = os.path.realpath(os.path.dirname(sys.argv[0]))
++
++config_filename = os.path.realpath(os.path.join(program_path, '../config.xml'))
++
++
++
++config = configxml.ConfigFile(config_filename).read()
++
++
++print config.config.__dict__.get(sys.argv[1])
++
+--- /dev/null
++++ boinc-7.0.38+dfsg/stripchart/samples/dir_size
+@@ -0,0 +1,8 @@
++#! /bin/csh
++
++set SIZE = `du -sk $1 | awk '{print $1*1024}'`
++
++set CIVDATE = `/bin/date "+%Y:%m:%d:%H:%M"`
++set UNIXDATE = `/usr/local/bin/perl -e 'print time()'`
++
++echo $CIVDATE $UNIXDATE $SIZE
+--- /dev/null
++++ boinc-7.0.38+dfsg/stripchart/samples/datafiles
+@@ -0,0 +1,15 @@
++# This is the list of data sources for stripchart, in the colon-delimited format:
++# path to data file:title for the graph:column number in the data file:flags to stripchart
++#
++# examples:
++# /disks/matt/data_files/connections:connection drops:3
++# /disks/matt/data_files/connections:tcp rate:4:-b 100
++#
++# To put in a horizontal rule for ease of selection, use the line:
++# x:----------------------------------:x
++#
++get_load_out:machine load:3
++count_results_out:Number of results:3
++assimilated_wus_out:Number of assimilated workunits:3
++download_size_out:Bytes in download dir:3
++upload_size_out:Bytes in upload dir:3
+--- /dev/null
++++ boinc-7.0.38+dfsg/stripchart/samples/db_looper
+@@ -0,0 +1,4 @@
++#! /bin/csh
++
++set ARG = "db_count $1"
++looper "$ARG" $2
diff --git a/debian/patches/series b/debian/patches/series
index 678daf2..3c89f68 100644
--- a/debian/patches/series
+++ b/debian/patches/series
@@ -69,3 +69,4 @@ dynamic_libboinc.patch
Missing_libSched.patch
missing_libs_wrapper.patch
add-missing-test-files.patch
+add-missing-stripchart-samples.patch
--
BOINC packaging
More information about the pkg-boinc-commits
mailing list