tags/lttv-0.11.3-23102008/doc/developer/lttng-lttv-roadmap.html

   1 <html>
   2 <body>
   3 <center><big><big>LTTV & LTTng roadmap<small><small></center>
   4 <br>
   5 <br>
   6 Here are the roadmaps for the LTTV and LTTng development. I use a priority indice
   7 for the TODO items :<br>
   8 (1) : very high priority<br>
   9 (10): lowest priority<br>
  10 <br>
  11 <br>
  12 Dependencies are written between brackets [ ].<br>
  13 The # symbol marks who is currently working on the item.<br>
  14 The % symbol marks who is interested in the realisation of the item.<br>
  15 <br>
  16 <big>LTTV Roadmap<small><br>
  17 <br>
  18 * TODO<br>
  19 (3) Add cluster support.<br>
  20         # <A HREF="mailto:Eric Clement <eric.clement@polymtl.ca>">Eric Clement</A><br>
  21 (3) Make LTTV aware of type formats (visual separators) defined in the XML
  22 file.<br>
  23         # <A HREF="mailto:Gabriel Matni <gabrien.matni@polymtl.ca>">Gabriel Matni</A><br>
  24 (4) Statistics per time window.<br>
  25 (4) Add Xen per physical CPU view.<br>
  26 (4) Add Xen per vcpu view.<br>
  27 (4) Disable plugins when threshold reached (i.e. too much process in control
  28 flow view). Draw, and, when the threshold is reached, stop drawing. The global
  29 statistics view can inhibit showing the per process stats.<br>
  30 (4) Add a visual artifact : PID 0 could be named swapper instead of UNNAMED for
  31 cpus > 0.<br>
  32 (4) Add event specific fields support to filter.<br>
  33 (4) Add a periodic event interval view. (useful to verify event periodicity)<br>
  34 (4) create a graphical per cpu activity view.<br>
  35 (4) Filter by target process.<br>
  36 (4) Compensate for time spent in probes in LTTV analysis.<br>
  37 (4) Add CPU, network, disk, memory usage histogram. [Per interval statistics]<br>
  38 (4) Add sort by process priority in the control flow view (must also instrument
  39 priority information of the processes).<br>
  40 % Airbus<br>
  41 (5) Add Python scripting hooks.<br>
  42 (5) Add GUI interface to take an hybrid trace.<br>
  43 (5) Flight recorder : start lttd automatically upon GUI trace control stop.<br>
  44 (5) Automatically detect traces with too much processes and disable faulty operations.<br>
  45 (5) Event sequence detector (inspired from regular expressions).<br>
  46 (7) Create a hardware counter viewer (low cost rate counters : L1 cache miss,
  47 page faults, interrupts...). This will be a generalisation of the event rate
  48 view into a view of the evolution of a user definable event field.<br>
  49 <br>
  50 * TO FIX<br>
  51 (10) Add cancel button to LTTV filter GUI window.<br>
  52 (10) Sometimes, in the control flow view, a process with 0 creation time is
  53 created in addition to the real process itself. Seems to be caused by end of
  54 process life.<br>
  55 (10) Statistics do not take in account the time spent in the mode present at
  56 the beginning of the trace. Example : real time spent in system call on behalf
  57 of process 0.<br>
  58 <br>
  59 <br>
  60 <big>LTT Next Generation Roadmap<small><br>
  61 <br>
  62 * TODO<br>
  63 (1) Dump mounts.<br>
  64 (1) Bring userspace tracing in sync with new lttng.<br>
  65 (1) Bring stack dump in sync with new lttng.<br>
  66 (1) efficient dynamic event filtering while recording trace.<br>
  67         % Google<br>
  68         % Sensis Corp. <A HREF="mailto:Bish, Tim <Tim.Bish@Sensis.com>">Tim Bish</A><br>
  69         # <A HREF="mailto:Mathieu Desnoyers <compudj@krystal.dyndns.org>">Mathieu Desnoyers</A><br>
  70         - Partly implemented: per marker selection through /proc interface.<br>
  71         - Must still implement per active trace filtering.<br>
  72 (1) Add Xen support. (Trace buffer desallocation needs to be fixed)<br>
  73         # <A HREF="mailto:Mathieu Desnoyers <compudj@krystal.dyndns.org>">Mathieu Desnoyers</A><br>
  74 (1) Integrate SystemTAP logging with LTTng.<br>
  75 (4) instrument kernel bottom half irqsave, seqlocks, semaphores, mutexes, brlock.<br>
  76 (4) integrate NPTL instrumentation (see
  77 <A HREF="http://nptltracetool.sourceforge.net/">PTT</A>).<br>
  78 (4) Probe calibration kernel module.<br>
  79 (4) Make page faults detect nested fault without nesting 4 times in the page
  80 fault handler.<br>
  81 (5) Support CPUs with scalable frequency.<br>
  82 (5) Add boot time tracing support.<br>
  83 (5) Integrate LTTng and lttd with LKCD.<br>
  84 (7) Integrate periodical dump of perfctr hardware counters.<br>
  85 (8) Integrate periodical dump of SystemTAP computed information.<br>
  86 (9) Add support for setjmp/longjmp and jump tables instrumentation to
  87 ltt-instrument-functions.<br>
  88 (9) Make ltt-usertrace-fast support internal heartbeat timer.<br>
  89 (3) port LTTng to :<br>
  90 alpha<BR>
  91 user-mode Linux<BR>
  92 Xen<BR>
  93 # IBM<BR>
  94 S/390<BR>
  95 RTLinux<BR>
  96 % Wind River for 2.6.14<BR>
  97 sparc64<BR>
  98 # Wind River<BR>
  99 sh4<br>
 100 <br>
 101 <br>
 102 <br>
 103 * TODO (low priority)<br>
 104 enhance RPM packages for lttng kernel<br>
 105 Integrate header generation (genevent) in kernel build system.<br>
 106 Export channels via network sockets instead of writing them to disk.<br>
 107 Export buffers with time constraint for "live" visualisation. Use
 108 ltt_force_switch periodically from a timer to insure slow channels do not
 109 interfere with viewing.<br>
 110 Have an optional round-robin mode to write information into multiple channels
 111 from the same source.<br>
 112 <br>
 113 * Need to be discussed<br>
 114 Use 2.6.14 RelayFS control files.<br>
 115 Drop ltt-module-register and ltt-module-unregister, use exported variables.<br>
 116 drop ltt_filter_control, use functions pointers instead.<br>
 117 Merge facilities headers into one big header.<br>
 118 Change the name of XML files from XML to something else.<br>
 119 Remove ltt-base.c.<br>
 120 <br>
 121 Mathieu Desnoyers<br>
 122
 123
 124 </body>
 125 </html>