Github Migration - Resources
by Stefan Negrea
Hello Everybody,
Thank you for participating in this morning's session about the github migration. Here are some resources to help with migration:
1) Fork a repo: https://help.github.com/articles/fork-a-repo
Step 3 talks about multiple remotes. In our case, the remotes are fedorahosted and github. The man page for git-remote is also helpful for this.
2) SSH Key Generation: https://help.github.com/articles/generating-ssh-keys
I recommend using -b flag with a value higher or equal 2048 and also use a password for the key. If you want to add a password to the key but not be prompted every time look into ssh-add (Step 2 briefly mentions it).
3) SSH Configuration: https://gist.github.com/jexchan/2351996
This gist shows a way to configure the ssh client to support multiple keys and host specific keys.
This morning I was running into an issue with the ssh client because I had ForwardAgent enabled. So the ssh client was picking up the wrong key chain from the host I originated the ssh connection chain.
Please let me know if you have any other questions or need help with the migration.
Thank you,
Stefan Negrea
Software Engineer
10 years, 2 months
Bug 1054360 - RFE: let users run an AS7/EAP6 CLI script with an operation on Standalone or Host Controller resources
by Libor Zoubek
Hello,
I started to work on https://bugzilla.redhat.com/show_bug.cgi?id=1054360 I
have several more or less cosmetic questions:
1. In order to run jboss-cli plugin needs to know AS7 native management
port and interface. I've introduced 1 pluginConfig property <host>:<port>
and added it into "operations" group. Could it be considered as correct? I
mean I could create 2 properties (similar to what we have for http
management interface host and port).
2. Plugin in the end runs jboss-cli process with several parameters. I am
setting ProcessExecution.setWaitForCompletion some hardcoded value to wait
for process to finish. Is it possible/recommended to fill the value with
timeout for operation defined in plugin descriptor? When timeout
ProcessExecution.setWaitForCompletion is reached, child process is killed,
I tend to set such timeout to very big number, since I don't want child
script to get interrupted in case user did not correctly set operation
timeout (he did not expect his script to run too long).
4. What would you suggest as default timeout for "executeCommands" and
"executeScript" operations? (Iam now on 60s)
5. As scripts are expected to do AS configuration changes, would you
recommend to requestAvailabilityCheck() and
requestDeferredChildResourcesDiscovery() after successfull script
execution?
Thank you
--
Libor Zoubek
10 years, 2 months
Inclusion of libsigar-arm?
by Heiko W.Rupp
Hey,
I'd like to include libsigar-arm in 4.10 so that you can run the rhq-agent out of the box
(see http://pilhuhn.blogspot.de/2014/02/running-rhq-agent-on-raspberry-pi.html )
As this is not in the sigar distro we use, I wonder what the best way is to achieve
this.
Simplest is for sure to make a directory
modules/enterprise/agent/extra, put that library into it and then have in ant-run.xml
inside the agent module an additional line
<copy file="${basedir}/extra/libsigar-arm-linux.so" todir="${agent.distro.dir}/lib"/>
to copy the library into the target.
In a mavenized environment, this is only the 2nd best option of course.
So what would work better?
Heiko
10 years, 2 months
Github Overview - Phase 1 - Github Migration
by Stefan Negrea
The following meeting has been modified:
Subject: Github Overview - Phase 1 - Github Migration
Organizer: "Stefan Negrea" <snegrea(a)redhat.com>
Location: Google Hangout
Time: Monday, February 24, 2014, 9:00:00 AM - 10:00:00 AM GMT -06:00 US/Canada Central [MODIFIED]
Invitees: rhq-devel(a)lists.fedorahosted.org
*~*~*~*~*~*~*~*~*~*
Hello Everybody,
Part of the Github migration plan is to have an overview of Github in phase 1. I plan on going over the basic settings required for the migration such as ssh keys, account setup, and required local changes. I will also show a couple of interesting Github features.
Please let me know if you have any specific topics or questions that you want covered and will be glad to add to the meeting.
I will send the hangout details shortly before the meeting on #rhq channel on freenode.
Thank you,
Stefan Negrea
10 years, 2 months
Github Overview - Phase 1 - Github Migration
by Stefan Negrea
The following is a new meeting request:
Subject: Github Overview - Phase 1 - Github Migration
Organizer: "Stefan Negrea" <snegrea(a)redhat.com>
Location: Google Hangout
Time: Monday, February 24, 2014, 9:00:00 PM - 10:00:00 PM GMT -06:00 US/Canada Central
Invitees: rhq-devel(a)lists.fedorahosted.org
*~*~*~*~*~*~*~*~*~*
Hello Everybody,
Part of the Github migration plan is to have an overview of Github in phase 1. I plan on going over the basic settings required for the migration such as ssh keys, account setup, and required local changes. I will also show a couple of interesting Github features.
Please let me know if you have any specific topics or questions that you want covered and will be glad to add to the meeting.
I will send the hangout details shortly before the meeting on #rhq channel on freenode.
Thank you,
Stefan Negrea
10 years, 2 months
Manual add from CLI lead to duplicate resources
by Thomas Segismont
Hi everyone,
So I'm playing with the attached script to manually import a few
thousands of "PortServices" [1].
From time to time, resources are duplicated in inventory (see the
attached screenshots).
I'll start to look at this tomorrow but wanted to share in case it rings
a bell to somebody.
Thanks,
Thomas
[1] this is to test the fix for Bug 738799 - deleting a group alertdef
containing >1000 member alertdefs fails with "SQLException: ORA-01795:
maximum number of expressions in a list is 1000" error
10 years, 2 months
Github Migration - Plan
by Stefan Negrea
Hello Everybody,
Tooling and infrastructure are very important to any open source community. As stated in "RHQ State of the Union" email we will migrate the source code repository to github. The migration will just move the code from one host to another; no structural changes for the project are planned at this time. The developer and user mailing lists for the project will remain on fedorahosted.org.
Phase 1 - 2/24/2014 - Preparation
Scope:
- Clean current repository prior to migration
- Have github accounts ready
Team (prior to deadline):
- Delete any unneeded personal and BZ related branches
- If you have a specific branch that you want to keep please email me prior to this deadline
- Make sure you have a github account to be used post migration
- Setup the github account with a new ssh key and configure ssh client
Me (on 2/24):
- Will remove branches that point to closed BZs on fedorahosted.org
- Schedule a Google Hangout to discuss the migration and showcase Github
Phase 2 - 2/26/2014 - Trial migration
Scope:
- Attempt a migration (not the final migration)
Team: N/A
Me:
- Remove Github repositories from the previous migration attempt
- Attempt the migration to github to check for duration and potential issues
- If any issues arise I will try to mitigate them before the next phase, if not enough time then delay next phase
Phase 3 - 02/28/2014 - Migration
Scope:
- Migrate to github
Team (after migration):
- Update the config URL to point to github.
Me:
- Disable write access to all current contributors to fedorahosted.org repo (with Mazz's help)
- Migrate repository
- Enable write access to the new
- Send out detailed information on how to reconfigure local repositories clones
- Reconfigure release tooling and CI jobs to point to the repository location
Please let me know if you have any concerns about this plan. I will keep the team informed throughout the process, so watch for emails and meeting invites.
Thank you,
Stefan Negrea
Software Engineer
10 years, 2 months
Integrating RHQ with Fabric8
by Lukas Krejci
The following is a new meeting request:
Subject: Integrating RHQ with Fabric8
Organizer: "Lukas Krejci" <lkrejci(a)redhat.com>
Time: Wednesday, February 19, 2014, 4:30:00 PM - 5:30:00 PM GMT +01:00 Belgrade, Bratislava, Budapest, Ljubljana, Prague
Invitees: rhq-devel(a)lists.fedorahosted.org
*~*~*~*~*~*~*~*~*~*
This presentation will discuss the possibilities of integration of provisioning between RHQ and Fabric8.
If you want to shape the future of RHQ's provisioning, don't hesitate to join us on a public hangout (details of which will follow).
The hangout of the meeting will be on air and the discussion will be possible through G+ and IRC: #rhq@freenode.
https://plus.google.com/events/c9c5j1cvg1e2g2pb8i98lkf20k8
10 years, 2 months
Various other bugs for 4.10
by Elias Ross
I'm not sure you can get these in time for 4.10, but wanted to see if
there was interest.
Bug 918207 - Uninventorying resources for agents, on unreachable
networks or waiting for registration, causes database hang
Fix is to set a timeout when calling the agent to uninventory.
This is related to the below. If the agent is running, but not
registered (the agent lost its token), currently RHQ will simply wait
a long time to get acknowledged.
Bug 987628 - Agent can block indefinitely at startup if server has an issue
I have a patch for this, but awaiting approval by Apple. Basically if
registration fails, the agent will just stay up forever and never
shutdown, which is not good.
Bug 1010485 - Apache plugin's WWWUtil should not follow HTTP redirects
If your root file redirects you to another site, not good. Patch inline.
10 years, 2 months