Author Topic: Invitation to discuss methods of clustering Zentyal  (Read 4548 times)

stuartiannaylor

  • Zen Monk
  • **
  • Posts: 67
  • Karma: +1/-6
    • View Profile
Re: Invitation to discuss methods of clustering Zentyal
« Reply #15 on: December 04, 2012, 12:51:37 pm »
Christian thanks, your bang on the button in my mind.

The only area I differ is that I think actually zentyal will fit a multiple server ditsributed mode.
Zentyal is a strange product in that its a product that controls the settings of other products.

For me its a common admin interface that is a familar friend for the sysadmin. Because of this it makes configuration tasks very easy.
I have set up ubuntu services the hard way via manuals and /etc config files. What has taken days and much head scratching can be accomplished with zentyal in minutes.
A complete server with most of the zentyal module offerings can be implemented from an iso to blank drives to a working server in 30 minutes. That is a huge ammount of value.

I feel btrfs will provide the snapshots, dynamic inode allocation, efficient incremental backups, background scrub processes, object level striping and mirroring.
I feel LXC can provide the isolation and distribution required and whilst linking at the network layer and providing control through this great web admin console we are all familiar with.

I know btrfs and lxc are still under heavy development and I am not saying they should be employed now. I feel that development should have brtfs and lxc in mind as without doubt they will become common technologies in majority usuage.

The monolithic all in one approach is similar to XP EOL yes many will argue that it is a valid and usefull platform and on the flip side with people for an eye to the future will believe its already dead in the water.

As a product and light weight admin console that works at a network level, Zentyal is cutting edge and its future is only dependent on direction.   
« Last Edit: December 04, 2012, 02:15:07 pm by stuartiannaylor »

stuartiannaylor

  • Zen Monk
  • **
  • Posts: 67
  • Karma: +1/-6
    • View Profile
Re: Invitation to discuss methods of clustering Zentyal
« Reply #16 on: December 04, 2012, 01:17:09 pm »
(1)
Zentyal is not designed to handle "multiple Zentyal box" management. in case of multiple Zentyal servers, there is no "service redundancy" but different services, per server.
- no DNS replication, zone transfer but local DNS
- one single mail service per mail domain
- LDAP replication still not flexible enough (+ internal LDAP for Samba)

Christian I don't think zentyal should be designed to handle "multiple Zentyal box" management. I think it is what it is and its a per server admin console that administers the server settings locally.
I do think that via distributed admin consoles of common familarity could become part and parcel of a multiple zentyal box scenario.


stuartiannaylor

  • Zen Monk
  • **
  • Posts: 67
  • Karma: +1/-6
    • View Profile
Re: Invitation to discuss methods of clustering Zentyal
« Reply #17 on: December 04, 2012, 01:34:24 pm »
Its a really good idea, I have been meaning to have a play with a Pi. I have been interested to see how bodhilinux runs on one and just a whim of an idea the cheapest thin client possible.
<snip>

I have run bodhi on a PI.  It runs very nice indeed.  You might also take a look at https://www.miniand.com/products/Hackberry%20A10%20Developer%20Board I have one of these also.

I have a current love affair with bodhilinux to me its the perfect thin-client. I have it running on some seriously old and slow desktops and its amazing how such a beautiful desktop can be so slim and functional.
My users love it and find it simple to use and its ubuntu base has that huge scope that canonical provide.

On this forum we have a complete section called Zentyal desktop which is totally fubar. It is completely reinventing the wheel and again comes from this monolithic thought path of zentyal providing everything.

It has as much worth as the activity in development and in the forum. Totally misguided iniative that limits choice to a proscribed solution.

Zentyal is a server product that should have modules that allow multiple common distro's be it ubuntu, bodhilinux, fedora or suse to connect and utilise services.
The idea of a dedicated server and desktop platform is so M$ and narrow minded I could spit.
« Last Edit: December 04, 2012, 02:09:55 pm by stuartiannaylor »

half_life

  • Bug Hunter
  • Zen Hero
  • *****
  • Posts: 867
  • Karma: +59/-0
    • View Profile
Re: Invitation to discuss methods of clustering Zentyal
« Reply #18 on: December 05, 2012, 01:18:21 am »
It has been a very long day.  I have not forgotten this but I will be taking the evening off.

stuartiannaylor

  • Zen Monk
  • **
  • Posts: 67
  • Karma: +1/-6
    • View Profile
Re: Invitation to discuss methods of clustering Zentyal
« Reply #19 on: December 05, 2012, 02:57:36 am »
Hey everybody I am so apologetic as I know, I have brought to much angst to the forum.
I am passionate and I believe in zentyal not being just a product, collaborative meritocracy, pushing for a fairer more efficient solution.
Getting it right.
Two guys on here that I hold in the utmost respect are Sam Graff and Christian.
There knowledge and out look is truly outstanding.
I am apologetic because I have totally enjoyed finding some real voices with knowledge and I know I have been pushing to hard.
For me open source and collaboration isn't just a technical geek thing its a modern morality.
Lol I know, how it sounds but hey, I spent a year in South Africa and to be honest it has changed my life and perspective.

I have been really critical about the project because I believe in it.

I am court tomorrow and I am going to tell them to F*** off. Can't think of a better way to do it.
Hit a wave of depression as the last two years haven't gone my way.
I have been certain that I believe in open source meritocracy and very much a open source left winger.
I am back purely to get the community centres I have been involved with using samba4 and not M$.

I had this grand plan of open source and providing a solution that doesn't entrap users into a product racket.
Its been a total failure and in my sphere of educating grannies, displaced kids and generally the under privileged I have been trapped.
I have provided several centres with total open source solutions and can't get past the barriers of council education.

The tutors have M$ qualifications and teach only M$ products, show people Itunes and state that they are expensive but don't show alternatives.

Anyway I have had my own problems and an avid open source fanatic for want of better words. So much so I have refused to use M$ as a server product. With XP all was well, Vista was a thankfully a disaster but win7 has been a killer.
I have a two contracts left and been pretty much breadline for two years now.

I have been preaching Zentyal 3 and I know it will get there. Thing is if it is going to get any professional kudos then you can't release a beta. A Beta it is, and has been for a while.

Its a superb product and I believe it will be, but you can't release and advertise a product as it has been in its current state.
It just loses faith and it can have huge impacts on those who expect a normal cycle of a release to work.
Its simple its either a working bug checked release or you state that the bleeding edge is so and use as so.

The amount of users that have run into problems with 3 and looked at zentyal as a flaky product and gone elsewhere is truly depressing. All was needed was more conversation to the community.

This is open-source and be open, say it is how it is. That is why it is more efficient, that is why your users benefit via contribution.
Organic meritocracies are not gay, religious tree hugging freaks. If you have enough experience in IT then nothing is black and white. Denial and not being open just wastes time and doesn't indicate to knowledge users where there skills might be beneficial.

The trade mark of zentyal alone if allowed to move in similar directions to products such as MySql could make the original crowd very wealthy. There are huge financial incentives to be reaped.

This will never happen by blind allegiance and drunken friendships. It isn't fair to the community for a few to party and dictate on there own ego's.

http://trac.zentyal.org/wiki/Structure/Governance


It is so obvious its not a meritocracy, firstly the community council is top heavy and consists mainly of commercial zentyal staff.
That is not a community council.
I did for a brief moment become a community council member and was just stunned by the ego of the council as for some reason they believe they know better than the users of the product.
Going back to Sam and Christian the community is in good hands.

Sorry Rob and Ishat, but wow guys look at everything you have forwarded, been involved with... Its so far from a meritocracy it is just deeply sad.

That is just my opinion, but if you are going to have a meritocracy as stated then do so. If this is going to be the control of a SADFL then please state so. Just don't pretend in the long term that it is something that it plainly isn't.

Why can't the community vote and provide indication to members of value. This doesn't mean it that its set in stone and always up to the SADFL to adjudicate.

So anyway I have gone bank, my probs. Court thing is a bit of a worry as I haven't a clue where my mind was at. Generally I apologise and to a few I do actually have a huge amount of respect for and best wishes.

Wish the project the best but by providing more discussion to the community and being transparent increases trust and participation. Stop the commercial denial as its losing trust.

Stuart.

Best wishes seriously.