hi,
was there some planned outage for today? Web frontend is down, osc can't connect, etc etc.
bye, MH
Am Samstag, 26. März 2011, 14:45:08 schrieb Mathias Homann:
hi,
was there some planned outage for today? Web frontend is down, osc can't connect, etc etc.
No, this is an unplanned outage of ichain proxy afaics.
Greetings, Stephan
On Sat, Mar 26, 2011 at 14:05, Stephan Kulow coolo@suse.de wrote:
Am Samstag, 26. März 2011, 14:45:08 schrieb Mathias Homann:
hi,
was there some planned outage for today? Web frontend is down, osc can't connect, etc etc.
No, this is an unplanned outage of ichain proxy afaics.
Greetings, Stephan
ETA for service restoration?
On Sa 26 Mär 2011 23:15:29 CET Robert Xu robxu9@gmail.com wrote:
was there some planned outage for today? Web frontend is down, osc can't connect, etc etc.
No, this is an unplanned outage of ichain proxy afaics.
Greetings, Stephan
ETA for service restoration?
I currently would expect Monday.
Sorry for this - but at the moment there's noting left we can do here from our side.
CU, Lars
On 27/03/11 10:55, Lars Vogdt wrote:
Sorry for this - but at the moment there's noting left we can do here from our side.
We could (and should) create some effort to get rid of this "wonder of technology".
On Mon, Mar 28, 2011 at 12:36:16AM +0200, Pavol Rusnak wrote:
On 27/03/11 10:55, Lars Vogdt wrote:
Sorry for this - but at the moment there's noting left we can do here from our side.
We could (and should) create some effort to get rid of this "wonder of technology".
It is on the "list of things to do if..."
On Sun, Mar 27, 2011 at 03:47:11PM -0700, Greg KH wrote:
On Mon, Mar 28, 2011 at 12:36:16AM +0200, Pavol Rusnak wrote:
On 27/03/11 10:55, Lars Vogdt wrote:
Sorry for this - but at the moment there's noting left we can do here from our side.
We could (and should) create some effort to get rid of this "wonder of technology".
It is on the "list of things to do if..."
Lars or someone, did you at least open a ticket?
Ciao, Marcus
Am Montag, 28. März 2011, 09:03:26 schrieb Marcus Meissner:
On Sun, Mar 27, 2011 at 03:47:11PM -0700, Greg KH wrote:
On Mon, Mar 28, 2011 at 12:36:16AM +0200, Pavol Rusnak wrote:
On 27/03/11 10:55, Lars Vogdt wrote:
Sorry for this - but at the moment there's noting left we can do here from our side.
We could (and should) create some effort to get rid of this "wonder of technology".
It is on the "list of things to do if..."
Lars or someone, did you at least open a ticket?
A ticket is running since saturday, just got it also flagged as pressing issue ...
Am Montag, 28. März 2011 schrieb Pavol Rusnak:
On 27/03/11 10:55, Lars Vogdt wrote:
Sorry for this - but at the moment there's noting left we can do here from our side.
We could (and should) create some effort to get rid of this "wonder of technology".
It's easy to pick on a single point of failure when it's down, but as a matter of fact the OBS has about 5 of them - replacing ichain with another single of point of failure won't really help weekend outages IMO.
Greetings, Stephan
Am 28.03.2011 09:57, schrieb Stephan Kulow:
Am Montag, 28. März 2011 schrieb Pavol Rusnak:
On 27/03/11 10:55, Lars Vogdt wrote:
Sorry for this - but at the moment there's noting left we can do here from our side.
We could (and should) create some effort to get rid of this "wonder of technology".
It's easy to pick on a single point of failure when it's down, but as a matter of fact the OBS has about 5 of them - replacing ichain with another single of point of failure won't really help weekend outages IMO.
Depending on what and where it's running it could help. If data center access is needed there is not much hope in improvement since only volunteers are working at the weekend. This outage affected releasing security updates for Evergreen quite a lot :-(
Wolfgang
Am Montag, 28. März 2011 schrieb Wolfgang Rosenauer:
Am 28.03.2011 09:57, schrieb Stephan Kulow:
Am Montag, 28. März 2011 schrieb Pavol Rusnak:
On 27/03/11 10:55, Lars Vogdt wrote:
Sorry for this - but at the moment there's noting left we can do here from our side.
We could (and should) create some effort to get rid of this "wonder of technology".
It's easy to pick on a single point of failure when it's down, but as a matter of fact the OBS has about 5 of them - replacing ichain with another single of point of failure won't really help weekend outages IMO.
Depending on what and where it's running it could help. If data center access is needed there is not much hope in improvement since only volunteers are working at the weekend. This outage affected releasing security updates for Evergreen quite a lot :-(
ichain is about user data - switching from proxy model to e.g. ldap would allow us to show better errors in case the user data is down - and continue with public operations, but without a way to authentificate yourself the OBS won't be able to security updates for Evergreen. That's what I'm saying.
Greetings, Stephan
On Mon, Mar 28, 2011 at 4:57 AM, Stephan Kulow coolo@suse.de wrote:
Am Montag, 28. März 2011 schrieb Pavol Rusnak:
On 27/03/11 10:55, Lars Vogdt wrote:
Sorry for this - but at the moment there's noting left we can do here from our side.
We could (and should) create some effort to get rid of this "wonder of technology".
It's easy to pick on a single point of failure when it's down, but as a matter of fact the OBS has about 5 of them - replacing ichain with another single of point of failure won't really help weekend outages IMO.
At work we had one such POF, there was a memory fragmentation bug that was very hard to track down, so in the meantime we placed a script that checked when the process ran rampant and restarted the service.
Dirty, but it did buy us some time to actually track and fix the bug.
Can something like that be done here?
On 28/03/11 09:57, Stephan Kulow wrote:
It's easy to pick on a single point of failure when it's down, but as a matter of fact the OBS has about 5 of them - replacing ichain with another single of point of failure won't really help weekend outages IMO.
Agreed, but mine point was that this PoF is not in our hands and we have to wait for someone who is not as passionate about openSUSE as we are ...
buildservice@lists.opensuse.org
-
Adrian Schröter -
Claudio Freire -
Greg KH -
Lars Vogdt -
Marcus Meissner -
Mathias Homann -
Pavol Rusnak -
Robert Xu -
Stephan Kulow -
Wolfgang Rosenauer