Good Morning All, There is lots of food for thought in this thread.
If beber does not have time and is willing to hand over total control to the infra to me I am more than willing to work on things as admin. I understand everyone wants to simplify things tremendously and I have to agree there. Also documentation is key here. And I will make sure everything gets documented for who ever joins me or even if i have to step down after me. Here is some food for thought. What about using a microservices management platform such as puppet for example to manage everything on the primary server as well as the secondary one in france? --- Regards, Jonathan Aquilina On 2017-06-09 10:17, Carsten Haitzler wrote: > On Fri, 09 Jun 2017 06:52:30 +0200 Jonathan Aquilina <jaquil...@eagleeyet.net> > said: > >> Carsten, I have expressed interest to help as an admin, but given the > > i know. but in this thread it's all "just move to the cloud and everything > will > be fine". it's a total distraction from the real issue. i've never seen this > kind of brokenness ever on linux, but i dont run custom kernels (haven't for > decades). fix the problem, don't avoid it as if "cloud is the answer". it's > not. > > complex undocumented nature of the current setup and beber needing me to > have a pgp which I only just got in all honesty Im scared to touch the > server. Like others have said we need to simplify things and get rid of > all this unnecessary complexity. Staying with VM's wont hurt anything. > For all we know even if we go to docker containers we could still end up > with the same instabilities like we are seeing with the current setup. > > --- > Regards, > > Jonathan Aquilina > > On 2017-06-09 02:15, Carsten Haitzler wrote: > > On Thu, 08 Jun 2017 18:17:32 +0000 Andrew Williams <a...@andywilliams.me> > said: > > Hi team, > > I get that the server itself is not problematic but can we please agree > that redundancy would be a good thing? > There is no way that we should be comfortable with a single issue be able > to take out our entire infrastructure. > redundancy (mirrored to another machine) is simply not possible. it would > require then we have to keep in sync multiple git servers, mysql db's and > that synchronisation would be broken when things go slow so any backup > would have different data and we'd have a nightmare to reconcile things. we > have full data redundancy at the filesystem/block layer (raid mirroring). > the entire infra is not down. the host is still running just fine (e5). > > if we move some services to new hosts, that means then you just have some > things broken. if git still runs on e5 then git is down. whatever runs there > will be down. moving everything is the same effort - in fact more, as fixing > things to move to containers. and it's now far riskier a it will become > expensive as people have to pay for that hosting. > > so NO. no moving to other systems. it's not a solution. it's sticking your > head in the sand and not resolving the issue we have. > > not a single person here ever says "hey - i'll help migrate it to a > container on e5 and figure it out". it's always "hey hey - come over here > and use the vm/container/whatever hosting provider i already use". > > e.org was stable as all hell when it was simple. it was a single machine > (with a spare we just never used because the main system never went down). > it had no vm's. no containers. it just ran everything on a single shared > system. it worked a charm. it had no performance issues (other than it > wasn't going to have the spare grunt to run a CI/build server e.g. > jenkins). our reason for a new system was to be able to do CI builds and > support and fear that our old machine would die (our drives were not raided > at all back then)... > > i'm also getting impatient. at e dev day in 2 weeks lets sort this out and > actually get something done there and then. tbh i think at this point this > is more important than just about any other topic there... and not to just > talk about, but to actually do immediately. if a few people sit around > converting what we have into something simpler (docker images? something > else?) and beber helps get it set up and goes over what is set up and why > so we all understand better how it's glued together... > > Thanks, > Andrew > > On Thu, 8 Jun 2017 at 06:39 Carsten Haitzler <ras...@rasterman.com> wrote: > > On Thu, 08 Jun 2017 14:35:39 +0200 Jonathan Aquilina < > jaquil...@eagleeyet.net> > said: > > Beber you have my pgp key through keybase.io's application even sent you > a chat message. Also I noticed you are coming to the dev summit in 2 > weeks time. Any chance we can sit down during the weekend and sort out > the server? > it's gotten far worse in recent times. like every day it needs a reboot > these > days... > > --- > Regards, > > Jonathan Aquilina > > On 2017-06-08 14:21, Amitesh Singh wrote: > > On Wed, Jun 7, 2017 at 11:25 PM, Mike Blumenkrantz < > michael.blumenkra...@gmail.com> wrote: > > the new error page is an improvement at least > Its down again :/ > > ------------------------------------------------------------ > ------------------ > Check out the vibrant tech community on one of the world's most > engaging tech sites, Slashdot.org! http://sdm.link/slashdot > _______________________________________________ > enlightenment-devel mailing list > enlightenment-devel@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/enlightenment-devel ------------------------------------------------------------------------------ >> Check out the vibrant tech community on one of the world's most >> engaging tech sites, Slashdot.org! http://sdm.link/slashdot >> _______________________________________________ >> enlightenment-devel mailing list >> enlightenment-devel@lists.sourceforge.net >> https://lists.sourceforge.net/lists/listinfo/enlightenment-devel ------------------------------------------------------------------------------ > Check out the vibrant tech community on one of the world's most > engaging tech sites, Slashdot.org! http://sdm.link/slashdot > _______________________________________________ > enlightenment-devel mailing list > enlightenment-devel@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/enlightenment-devel -- ------------- Codito, ergo sum - "I code, therefore I am" -------------- The Rasterman (Carsten Haitzler) ras...@rasterman.com ------------------------------------------------------------------------------ Check out the vibrant tech community on one of the world's most engaging tech sites, Slashdot.org! http://sdm.link/slashdot _______________________________________________ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel -- http://andywilliams.me http://ajwillia.ms ------------------------------------------------------------------------------ Check out the vibrant tech community on one of the world's most engaging tech sites, Slashdot.org! http://sdm.link/slashdot _______________________________________________ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel ------------------------------------------------------------------------------ Check out the vibrant tech community on one of the world's most engaging tech sites, Slashdot.org! http://sdm.link/slashdot _______________________________________________ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
