Client Tools for UNIX
What You'll Need
- A Backtrace account (log in or sign up for a free trial license).
- Your subdomain name (used to connect to your Backtrace instance). For example,
https://example-subdomain.sp.backtrace.io
. - A Backtrace project and a submission token.
Client Tools Installation
These steps require a license number from Backtrace to install the necessary packages. Contact us to receive a license number before proceeding.
All tools are installed in /opt/backtrace/bin/
by default. We recommend you add this to your path:
$ export PATH=$PATH:/opt/backtrace/bin
Coroner
coroner
is the Backtrace object store client. It is used to submit and query the object store. Check out the coroner client usage for more information.
RHEL/CentOS
$ yum install backtrace-coroner
Ubuntu/Debian
$ apt-get install backtrace-coroner
After installed, you need to generate and install coroner.cf.
ptrace
ptrace
is the Backtrace snapshot generator. It is used to generate application snapshots.
RHEL/CentOS
$ yum install backtrace-ptrace
Ubuntu/Debian
$ apt-get install backtrace-ptrace
Starting from Ubuntu 10.10, a change prevents users who are not root from ptrace
processes that are not children. With this restriction in place, running ptrace
as root is necessary, resulting in the snapshot files being permissioned for root as well.
This restriction can be disabled by issuing the command:
$ echo 0 | sudo tee /proc/sys/kernel/yama/ptrace_scope
For additional information, refer to the following comprehensive guide: After Upgrade, GDB Won't Attach to Process on the Ask Ubuntu website.
Hydra
hydra
is the Backtrace terminal snapshot viewer. It is used to dig into application snapshots generated by backtrace-ptrace
. You must install hydra
on any machine you wish to dig into snapshots (for example, dev box, laptop).
RHEL/CentOS
$ yum install backtrace-hydra
Ubuntu/Debian
$ apt-get install backtrace-hydra
Once installed, you may want to configure Hydra.