From Fedora Project Wiki
(make information up to date)
 
(14 intermediate revisions by 5 users not shown)
Line 1: Line 1:
= Retrace Server =
= Retrace Server =
== Summary ==
== Summary ==
Server that simplifies coredump analysis.
Retrace server allows ABRT users to get better backtraces from their crashes by retracing coredumps remotely, on a server owned by Fedora Project.


== Owner ==
== Owner ==
Line 12: Line 12:
=== Co-workers ===
=== Co-workers ===
* The ABRT team
* The ABRT team
* E-mail: crash-catcher@lists.fedoraproject.org
* Name: [[User:kklic|Karel Klic]]
* Name: [[User:kklic|Karel Klic]]
* Name: [[User:jmoskovc|Jiri Moskovcak]]
* Name: [[User:jmoskovc|Jiri Moskovcak]]
Line 17: Line 18:
== Current Status ==
== Current Status ==
* Targeted release: [[Releases/15 | Fedora 15]]
* Targeted release: [[Releases/15 | Fedora 15]]
* Last updated: 2010-11-09
* Last updated: 2011-05-13
* Percentage of completion: 50%
* Percentage of completion: 100%


== Detailed description ==
== Detailed description ==
At the moment of backtrace creation user needs to have all relevant debuginfos installed locally. Downloading and installing them may require a lot of time and storage space depending on the package and its dependencies. The idea of Retrace Server is to liberate users from this: User sends the coredump packed along with some other required information (e.g. packages list) to the Retrace Server, the Retrace Server installs required debuginfo files, creates backtrace using standard debugging tools and provides it to the user for further processing.
When ABRT generates a backtrace from a coredump, it needs debuginfo data to be available for the binary and all libraries involved in the crash. Debuginfo packages require a lot of storage space, and sometimes they are not available at all -- a package update causes the removal of older updates of the same package and their debuginfo.
 
Another problem is that GDB (which generates the backtrace from the coredump) needs data from the binary and libraries that were involved in the crash. If a user updated some relevant package between observing the crash and reporting it, he might be unable to generate a good quality backtrace because of the updates. This happens often, because Fedora is updated frequently.
 
Retrace server is one possibility to solve these issues. ABRT offers user to upload her coredump to a remote server, then the retracing step happens there. The server creates an environment identical to what was on the user's computer at the time of the crash, by installing all the required packages and their debuginfo. The retrace server is able to do that because it keeps all the older packages from updates, and relevant part of updates-testing locally on the server.
 
After creating the backtrace, only the submitter is allowed to download/view it.


== Benefit to Fedora ==
== Benefit to Fedora ==
No need to download debuginfo packages while creating backtrace from various binary crashes.
Users:
# Less disk space and processing time needed to use ABRT to report crashes.
# Possibility to report older crashes.
# Lower chance of failure, less time spent on crashes which cannot be retraced because debuginfo is no longer available.
 
Developers:
# Higher quality of ABRT reports.
# Possibility to quickly get a backtrace from any/random Fedora coredump.


== Scope ==
== Scope ==
The project itself consists of 3 parts:
ABRT is extended to support Retrace Server.
* '''Interface''': Receives the archive from user, unpacks all files, puts new task into the queue.
* '''Analyzer''': Takes the task from queue, for each creates a virtual root with all required packages and debuginfos installed and creates the backtrace within.
* '''Repo Synchronizer''': Downloads packages to a local repository containing all versions of all packages, to be able to handle crashes from older versions.


[[Features/ABRT|ABRT]] plugin that packs required files and sends the archive to the Retrace Server must be implemented.
Retrace server is installed on https://retrace.fedoraproject.org.
 
The server implementation consists of three parts:
* '''HTTP Interface''': Receives the archive with a coredump from user, unpacks all files, puts new task into the queue.
* '''Analyzer''': Takes a task from queue, creates a virtual root with all required packages and debuginfos installed and runs GDB to create the backtrace.
* '''Repository Synchronizer''': Downloads packages to a local repository containing all versions of all packages (no removal of older updates).


== How To Test ==
== How To Test ==
The communication between client and server will be handled by future ABRT plugin. As soon as it is implemented, the testing script will be removed.
CLI retrace client can be built from [http://git.fedorahosted.org/git/?p=abrt.git ABRT Git repository]. CLI is integrated into latest ABRT GUI (analyze_RetraceServer).


'''Usage'''


At the moment, testing is possible using "test-uploader" from "retrace" branch of [http://git.fedorahosted.org/git/?p=abrt.git ABRT Git repository]. The script is only designed for test purposes and does not handle all errors possible.
<pre>
$ abrt-retrace-client
Usage: abrt-retrace-client <operation> [options]
Operations: create/status/backtrace/log/batch


'''Usage:'''
    -v, --verbose        Be verbose
    -s, --syslog          log to syslog
    -k, --insecure        allow insecure connection to retrace server
    --url URL            retrace server URL
    --headers            (debug) show received HTTP headers


The script ''upload.py'' takes two arguments:
For create and batch operations
    -d, --dir DIR        read data from ABRT crash dump directory
    -c, --core COREDUMP  read data from coredump (not yet implemented)
    --no-unlink          (debug) do not delete temporary archive created from dump dir in /tmp


First (mandatory) - ABRT crash directory, by default found in the /var/spool/abrt/ directory.
For status, backtrace, and log operations
    -t, --task ID        id of your task on server
    -p, --password PWD    password of your task on server
</pre>


Second (optional) - Retrace server address (not important at the moment, there is only one testing machine running retrace server - simona.expresmu.sk - will be migrated).
Create a new retrace job:
<pre>
$ abrt-retrace-client create -d ~/.abrt/spool/ccpp-whatever
Task ID: 123456789
Task Password: qwerty123456
</pre>


Running ''upload.py'' with different number of arguments will display short help message.
Ask for status:
<pre>
$ abrt-retrace-client status -t 123456789 -p qwerty123456
PENDING / FINISHED_SUCCESS / FINISHED_FAILURE
</pre>


Ask for log:
<pre>
$ abrt-retrace-client log -t 123456789 -p qwerty123456
log (text/plain)
</pre>


Script's output includes raw HTTP response containig ''X-Task-Id'' and ''X-Task-Password'' headers. You may ask the retrace server about three things using these headers:
Ask for backtrace:
* '''Status''' - HTTP response contains ''X-Task-Status'' header with one of three values 'PENDING', 'FINISHED_SUCCESS', 'FINISHED_FAILURE'. wget may be used to show the output (the HTTP response has no body):
<pre>
<nowiki>wget -S --no-check-certificate -O /dev/null --header="X-Task-Password: <Task-Password>" "https://simona.expresmu.sk/<Task-Id>/"</nowiki>
$ abrt-retrace-client backtrace -t 123456789 -p qwerty123456
* '''Log''' - Afrer retrace is finished (FINISHED_SUCCESS or FINISHED_FAILURE) the log is available. wget may be used to download the log:
backtrace (text/plain)
<nowiki>wget -S --no-check-certificate --header="X-Task-Password: <Task-Password>" "https://simona.expresmu.sk/<Task-Id>/log"</nowiki>
</pre>
* '''Backtrace''' - Afrer successful retrace (FINISHED_SUCCESS) the backtrace is available. wget may be used to download the backtrace:
<nowiki>wget -S --no-check-certificate --header="X-Task-Password: <Task-Password>" "https://simona.expresmu.sk/<Task-Id>/backtrace"</nowiki>


Batch processing (all-in-one):
<pre>
$ abrt-retrace-client batch -d ~/.abrt/spool/ccpp-whatever
[create]
[status] PENDING
[status] PENDING
[status] PENDING
[status] FINISHED_SUCCESS
[backtrace]
</pre>


Each Task-Password is for single use. After every status / log / backtrace request, the response contains ''X-Task-Password'' header with new Task-Password.
The retrace server at {{filename|retrace01.fedoraproject.org}} (used by default) is configured to handle crashes for the following releases:
* Fedora 15 i686
* Fedora 15 x86_64


== User Experience ==
== User Experience ==
If ABRT catches a crash that needs to be retraced, user will be able to choose whether to generate backtrace locally or remotely.
ABRT reporting wizard offers a possibility to use retrace server instead of local retracing to generate a backtrace. Local retracing is still the default action.


== Dependencies ==
== Dependencies ==
* abrt-addon-ccpp
* new version of {{package|abrt|ABRT}}, supporting the retrace server option
* gdb
* a server with the retrace server application up and running
* xz
 
* mod_wsgi
Server code dependencies:
* python-webob
* {{package|mock}}
* {{package|xz}}
* {{package|elfutils}}
* {{package|createrepo}}
* {{package|mod_ssl}}
* {{package|mod_wsgi}}
* {{package|python-webob}}


== Contingency Plan ==
== Contingency Plan ==
Not required, feature can easily be disabled.
Hide the possibility of using the retrace server from users in ABRT's graphical user interface.


== Documentation ==
== Documentation ==
* https://fedorahosted.org/abrt/
* [https://fedorahosted.org/abrt/attachment/wiki/AbrtRetraceServer/retrace-server Design document]
* [https://fedorahosted.org/abrt/wiki/AbrtRetraceServer Retrace server wiki page on ABRT wiki]


== Release Notes ==
== Release Notes ==
* '''FIXME'''
ABRT, a crash reporting tool in Fedora, now allows to prepare a part of crash processing remotely, on a server owned by Fedora Project. Remote coredump retracing leads to better quality of reports. Retrace server can generate good backtraces with much higher success rate than local retracing.


== Comments and Discussion ==
== Comments and Discussion ==
* See [[Talk:Features/RetraceServer]]
* See [[Talk:Features/RetraceServer]]


[[Category:FeatureReadyForWrangler]]
[[Category:FeatureAcceptedF15]]

Latest revision as of 12:10, 13 May 2011

Retrace Server

Summary

Retrace server allows ABRT users to get better backtraces from their crashes by retracing coredumps remotely, on a server owned by Fedora Project.

Owner

  • Email: mtoman@redhat.com

Co-workers

Current Status

  • Targeted release: Fedora 15
  • Last updated: 2011-05-13
  • Percentage of completion: 100%

Detailed description

When ABRT generates a backtrace from a coredump, it needs debuginfo data to be available for the binary and all libraries involved in the crash. Debuginfo packages require a lot of storage space, and sometimes they are not available at all -- a package update causes the removal of older updates of the same package and their debuginfo.

Another problem is that GDB (which generates the backtrace from the coredump) needs data from the binary and libraries that were involved in the crash. If a user updated some relevant package between observing the crash and reporting it, he might be unable to generate a good quality backtrace because of the updates. This happens often, because Fedora is updated frequently.

Retrace server is one possibility to solve these issues. ABRT offers user to upload her coredump to a remote server, then the retracing step happens there. The server creates an environment identical to what was on the user's computer at the time of the crash, by installing all the required packages and their debuginfo. The retrace server is able to do that because it keeps all the older packages from updates, and relevant part of updates-testing locally on the server.

After creating the backtrace, only the submitter is allowed to download/view it.

Benefit to Fedora

Users:

  1. Less disk space and processing time needed to use ABRT to report crashes.
  2. Possibility to report older crashes.
  3. Lower chance of failure, less time spent on crashes which cannot be retraced because debuginfo is no longer available.

Developers:

  1. Higher quality of ABRT reports.
  2. Possibility to quickly get a backtrace from any/random Fedora coredump.

Scope

ABRT is extended to support Retrace Server.

Retrace server is installed on https://retrace.fedoraproject.org.

The server implementation consists of three parts:

  • HTTP Interface: Receives the archive with a coredump from user, unpacks all files, puts new task into the queue.
  • Analyzer: Takes a task from queue, creates a virtual root with all required packages and debuginfos installed and runs GDB to create the backtrace.
  • Repository Synchronizer: Downloads packages to a local repository containing all versions of all packages (no removal of older updates).

How To Test

CLI retrace client can be built from ABRT Git repository. CLI is integrated into latest ABRT GUI (analyze_RetraceServer).

Usage

$ abrt-retrace-client
Usage: abrt-retrace-client <operation> [options]
Operations: create/status/backtrace/log/batch

    -v, --verbose         Be verbose
    -s, --syslog          log to syslog
    -k, --insecure        allow insecure connection to retrace server
    --url URL             retrace server URL
    --headers             (debug) show received HTTP headers

For create and batch operations
    -d, --dir DIR         read data from ABRT crash dump directory
    -c, --core COREDUMP   read data from coredump (not yet implemented)
    --no-unlink           (debug) do not delete temporary archive created from dump dir in /tmp

For status, backtrace, and log operations
    -t, --task ID         id of your task on server
    -p, --password PWD    password of your task on server

Create a new retrace job:

$ abrt-retrace-client create -d ~/.abrt/spool/ccpp-whatever
Task ID: 123456789
Task Password: qwerty123456

Ask for status:

$ abrt-retrace-client status -t 123456789 -p qwerty123456
PENDING / FINISHED_SUCCESS / FINISHED_FAILURE

Ask for log:

$ abrt-retrace-client log -t 123456789 -p qwerty123456
log (text/plain)

Ask for backtrace:

$ abrt-retrace-client backtrace -t 123456789 -p qwerty123456
backtrace (text/plain)

Batch processing (all-in-one):

$ abrt-retrace-client batch -d ~/.abrt/spool/ccpp-whatever
[create]
[status] PENDING
[status] PENDING
[status] PENDING
[status] FINISHED_SUCCESS
[backtrace]

The retrace server at retrace01.fedoraproject.org (used by default) is configured to handle crashes for the following releases:

  • Fedora 15 i686
  • Fedora 15 x86_64

User Experience

ABRT reporting wizard offers a possibility to use retrace server instead of local retracing to generate a backtrace. Local retracing is still the default action.

Dependencies

  • new version of Package-x-generic-16.pngABRT, supporting the retrace server option
  • a server with the retrace server application up and running

Server code dependencies:

Contingency Plan

Hide the possibility of using the retrace server from users in ABRT's graphical user interface.

Documentation

Release Notes

ABRT, a crash reporting tool in Fedora, now allows to prepare a part of crash processing remotely, on a server owned by Fedora Project. Remote coredump retracing leads to better quality of reports. Retrace server can generate good backtraces with much higher success rate than local retracing.

Comments and Discussion