Re: [Spacewalk-devel] [RFC] showing timestamps in Spacewalk
- I do like the screen shots and idea of making date/time stuff easier to read - I've also seen people wanting date formatting displayed changed. Just because their language is English, doesn't mean that 11/12/13 is clear to them - since in America is is 12th November and in UK it is 11th December. +1 And this is the main reason, why I'd vote to have it configurable. I know Duncan does not like it, but it was discussed many times among Spacewalk developers and we agreed it really makes sense. Some users would like to display the dates according to their locales and some not. 11/12/13 is a wonderful example that not all the date formats are really usable, as it takes me half a minute to find out, what format is used from the other dates. E.g. As Spacewalk is not translated to my language, I use the English language preference, but do not like the slashes / within the date. My preferable date format is 17.12.13, sometimes 17.12.2013, but 2013-12-17 is also acceptable for me (but 13-12-17 definitely not!) Right, what fedora actually offers me for LC_TIME=cs_CZ.utf8 is Út pro 17, 09:30, what I really do not like. Now, what would you recommend me as an user? :-D Ok, and I forgot to mention, my colleague sitting next to me has a different preference. :-) What I think to be great to have are options: * timestamp according to locales * 2013-12-31 as this is a format, that everybody understands * simple parsing of d(d) m(m) yy(yy), their order and separator I know I wrote about the date format only, but time is much much easier. :-) Regards, -- Tomas Lestach Red Hat Satellite Engineering, Red Hat ___ Spacewalk-devel mailing list Spacewalk-devel@redhat.com https://www.redhat.com/mailman/listinfo/spacewalk-devel
Re: [Spacewalk-devel] [Patch] Client registration via proxy crashes
On 17/12/13 10:17, Flavio Castelli wrote: Fixed an error raised during the client registration via proxy when 'X-RHN-IP-Path' header is not set. This regression has been introduced by commit 251005196b02b. I found another issue later in the same part of the code. If you want I can merge the two patches into a single one. Cheers Flavio -- Flavio Castelli SUSE LINUX Products GmbH Maxfeldstraße 5, 90409 Nürnberg Germany From bc60a8ac7cfee9c3703ca52a629bf2a177830db4 Mon Sep 17 00:00:00 2001 From: Flavio Castelli fcaste...@suse.com Date: Tue, 17 Dec 2013 09:53:54 +0100 Subject: [PATCH] Fixed client registration via proxy The previous code relied on a wrong behavior of UserDict class which has been fixed with commit 251005196b02b. Fixed the code to not crash when the 'X-RHN-IP-Path' header is not set. --- proxy/proxy/broker/rhnBroker.py | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/proxy/proxy/broker/rhnBroker.py b/proxy/proxy/broker/rhnBroker.py index e77ef13..a9194ca 100644 --- a/proxy/proxy/broker/rhnBroker.py +++ b/proxy/proxy/broker/rhnBroker.py @@ -175,7 +175,9 @@ class BrokerHandler(SharedHandler): # do that for us # Add/modify the X-RHN-IP-Path header. -ip_path = _oto['X-RHN-IP-Path'] +ip_path = None +if 'X-RHN-IP-Path' in _oto: +ip_path = _oto['X-RHN-IP-Path'] log_debug(4, X-RHN-IP-Path is: %s % repr(ip_path)) client_ip = self.req.connection.remote_ip if ip_path is None: -- 1.8.4 ___ Spacewalk-devel mailing list Spacewalk-devel@redhat.com https://www.redhat.com/mailman/listinfo/spacewalk-devel
Re: [Spacewalk-devel] [RFC] showing timestamps in Spacewalk
On Tue, Dec 17, 2013 at 03:46:22AM -0500, Tomas Lestach wrote: What I think to be great to have are options: * timestamp according to locales * 2013-12-31 as this is a format, that everybody understands * simple parsing of d(d) m(m) yy(yy), their order and separator Don't forget am/pm vs. 24 hr thing. I agree. Support the default locale thing plus a bunch of reasonable formats to choose from. -- Jan Pazdziora Principal Software Engineer, Identity Management Engineering, Red Hat ___ Spacewalk-devel mailing list Spacewalk-devel@redhat.com https://www.redhat.com/mailman/listinfo/spacewalk-devel
Re: [Spacewalk-devel] [Patch] Client registration via proxy crashes
Hi Flavio, Unless I'm mistaken the patch attached to your second email is the same as the patch in your first. Want to send me the second one before I commit it? -Stephen On 12/17/2013 09:12 AM, Flavio Castelli wrote: On 17/12/13 10:17, Flavio Castelli wrote: Fixed an error raised during the client registration via proxy when 'X-RHN-IP-Path' header is not set. This regression has been introduced by commit 251005196b02b. I found another issue later in the same part of the code. If you want I can merge the two patches into a single one. Cheers Flavio ___ Spacewalk-devel mailing list Spacewalk-devel@redhat.com https://www.redhat.com/mailman/listinfo/spacewalk-devel ___ Spacewalk-devel mailing list Spacewalk-devel@redhat.com https://www.redhat.com/mailman/listinfo/spacewalk-devel
Re: [Spacewalk-devel] RFC: manage bare-metal systems from Spacewalk
Hey Silvio, this is an interesting feature. However I see the patches are applied on top of the power management feature and we haven't decided yet, whether to accept this feature. Are these patches independent from the power management feature, so I could apply them right now? I see a danger when looking at the changes in jsp and jspf files. Now, during the bootstrap integration, when the jsps get changed every day, we'll get at lot of merge conflicts. :( Regards, -- Tomas Lestach Red Hat Satellite Engineering, Red Hat - Original Message - From: Silvio Moioli smoi...@suse.de To: spacewalk-devel@redhat.com Sent: Friday, December 6, 2013 6:15:31 PM Subject: [Spacewalk-devel] RFC: manage bare-metal systems from Spacewalk Dear Spacewalk Team, as announced, this is the second part of the contribution we wanted to share with you. Actually it is again split in two sub-parts which are largely independent from one another and from the power management feature I wrote about yesterday - git commits are stacked over previous ones mostly for convenience reasons. First is a little addition to the SSM provisioning page - basically extending functionality provided by the Create Cobbler System Record button in the Provisioning page to multiple systems. https://git.fedorahosted.org/cgit/spacewalk.git/log/?h=master-create-cobbler-record-ssm As a bonus, these commits will also make provisioning errors visible in the SSM status bubble (they were basically ignored before). Second is support to list bare-metal systems in Spacewalk and provision them. This is the basic workflow: - admin enables this feature from Spacewalk Configuration - Bare-metal systems. This results in a new default Cobbler profile being created; - bare-metal system is powered on, and gets the default (hereinafter bootstrap) image from Cobbler. This is a special, minimal OS that will boot, register the system with a special bootstrap activation key and shut the system down; - admin will happily see its system in the Systems list with a special icon and slightly different pages (basically, to hide a lot of actions which would not make sense). Bare-metal systems have a special entitlement which gets assigned with the bootstrap activation key only; - provisioning works as usual with Kickstart by creating Cobbler records (obviously scheduled provisioning is not supported). Admin can also use the new feature above to provision multiple bare-metal systems; - admin can also use the power management feature to power on his system and begin the kickstart process. Code to implement the above is available here: https://git.fedorahosted.org/cgit/spacewalk.git/log/?h=master-pxe-default-image That basically comprises the following: - handling of bootstrap activation keys and entitlements, database updates; - new acls and front-end code to hide portions of the UI that do not make sense for bare-metal systems; - a new button to enable or disable this feature; - correct handling of a few corner cases, like migrating a bare-metal system between organizations; - test code. You may notice a big missing piece: the bootstrap image itself! We created one for SUSE Manager using SLES, and one for Spacewalk based on openSUSE. It is available, together with its sources, here: https://build.opensuse.org/package/show/systemsmanagement:spacewalk:misc/pxe-default-image As we lack knowledge in Fedora and RHEL tools to build images, we used Kiwi and OBS to build it, as we are mostly familiar with those - others in the Spacewalk community are of course welcome to produce a Fedora or RHEL-based version. Minimal end-user documentation for SUSE Manager, which is mostly relevant for Spacewalk as well, can be found here: http://wiki.novell.com/index.php/SUSE_Manager/NewFeatures2.1#Unattended_bare-metal_system_provisioning Some implementation details you might be interested in: - you can install a copy of the bootstrap image with the following commands. The last command is needed to allow Cobbler to read the image through SELinux: wget http://download.opensuse.org/repositories/systemsmanagement:/spacewalk:/misc/openSUSE_13.1/noarch/pxe-default-image-13.1-0.14.2.noarch.rpm rpm -i pxe-default-image-13.1-0.14.2.noarch.rpm chcon -R -v --type=cobbler_var_lib_t /srv/pxe-default-image - this feature needs a few new configuration options, particularly regarding the bootstrap image location (we kept it as a parameter to easily adapt to various distros). Copy those from rhn_java.conf or from the Troubleshooting section of the above user guide; - stored procedures have been updated to automatically add a large number of bootstrap entitlements to all organizations. That way system migrations between organizations are always possible; - enabling or disabling the feature also implicitly triggers a cobbler sync. On my standard Fedora 19 install, this
Re: [Spacewalk-devel] [Patch] Client registration via proxy crashes
These changes all look reasonable, committed as 70e86d8d47cc87bf7e2335489f4735e01bd90341 Thanks! -Stephen On 12/17/2013 10:41 AM, Flavio Castelli wrote: On 17/12/13 15:47, Stephen Herr wrote: Unless I'm mistaken the patch attached to your second email is the same as the patch in your first. Want to send me the second one before I commit it? You are right. That's the right patch I indented to send. Cheers Flavio ___ Spacewalk-devel mailing list Spacewalk-devel@redhat.com https://www.redhat.com/mailman/listinfo/spacewalk-devel
Re: [Spacewalk-devel] Asking for date/time in Spacewalk
On 17/12/13 08:53, Tomas Lestach wrote: Hey Duncan, the date picker is closely related with the displaying of timestamps as you among other things also display the proposed/chosen timestamp. I'd probably finish the displaying of timestamps and then concentrate on date picker. I think they are independent patches because we already have a date/time picker in spacewalk. It is worth noting that this implementation is backward compatible, even if it has 2 fields instead of 5 (?), it generates hidden compatibility form inputs for the old actions. I am posting them at the same time to start a discussion. I basically like the possibility to pick a date from the calendar and be able to enter it form the keyboard at the same time. Btw. I saw your patch requires jQuery. As I am not aware of any jquery packages available in Fedora koji http://koji.fedoraproject.org/koji/ the eventual patch would need to contain also packaging for all Fedora and RHEL versions for those Spacewalk will be built. jQuery is already on master. It was committed together with the UI refactoring as a replacement to prototype.js, which btw, was not packaged either. I know you would prefer to have jquery packaged, but the web-assets specification is pretty new (Fedora 20, http://fedoraproject.org/wiki/Changes/Web_Assets ), and you would had the same problem with the old prototype.js. Cheers -- Duncan Mac-Vicar P. - http://www.suse.com/ SUSE LINUX Products GmbH, GF: Jeff Hawn, Jennifer Guild, Felix Imendörffer, HRB 16746 (AG Nürnberg) Maxfeldstraße 5, 90409 Nürnberg, Germany ___ Spacewalk-devel mailing list Spacewalk-devel@redhat.com https://www.redhat.com/mailman/listinfo/spacewalk-devel