Server downtime on Saturday

[Colin Finck]

Hi all,

In the process of improving the performance of our infrastructure, we
will upgrade the first of two servers this Saturday.
This will result in a downtime for the following services:

* SVN
* Mail and mailing lists
* BuildBot
* Testman
* GIT mirror
* File server at svn.reactos.org
* Compatibility Database
* ev.reactos.org
* Second nameserver
* RosKGB bot

The upgrade will in particular increase CPU performance as well as RAM
and HDD sizes, so we will have more resources available afterwards.
Our current server has served us well since 2009, so don't expect such
an upgrade to happen again in the next few years.

We're aware that this affects many core services, so we will do our best
to keep the downtime as low as possible. Our target is <12h.
Gladly, we have already done such an upgrade in the past, so it mainly
boils down to the speed of transferring files over the network.
Meanwhile, we apologize for the caused inconveniences.

With best regards and on behalf of the whole infrastructure team,

Colin Finck

[Hermès BÉLUSCA - MAÏTO]

Ah :) This is nice !
By the way, I have a question: would it be useful to put such announcements
in some section on our website ?

Cheers,
Hermès.

-----Message d'origine-----
De : [e-mail] [mailto:[e-mail]] De la
part de Colin Finck
Envoyé : jeudi 21 novembre 2013 11:09
À : [e-mail]; ReactOS Development List
Objet : [ros-dev] Server downtime on Saturday

Hi all,

In the process of improving the performance of our infrastructure, we will
upgrade the first of two servers this Saturday.
This will result in a downtime for the following services:

* SVN
* Mail and mailing lists
* BuildBot
* Testman
* GIT mirror
* File server at svn.reactos.org
* Compatibility Database
* ev.reactos.org
* Second nameserver
* RosKGB bot

The upgrade will in particular increase CPU performance as well as RAM and
HDD sizes, so we will have more resources available afterwards.
Our current server has served us well since 2009, so don't expect such an
upgrade to happen again in the next few years.

We're aware that this affects many core services, so we will do our best to
keep the downtime as low as possible. Our target is <12h.
Gladly, we have already done such an upgrade in the past, so it mainly boils
down to the speed of transferring files over the network.
Meanwhile, we apologize for the caused inconveniences.

With best regards and on behalf of the whole infrastructure team,

Colin Finck

_______________________________________________
Ros-dev mailing list
[e-mail]
http://www.reactos.org/mailman/listinfo/ros-dev

[Colin Finck]

Hermès BÉLUSCA - MAÏTO [e-mail] wrote:
> Ah :) This is nice !
> By the way, I have a question: would it be useful to put such announcements
> in some section on our website ?

I wouldn't refrain from doing so if a website maintainer came to me and
told me where to put such an announcement :)

- Colin

[Zachary Gorden]

News or blog, pick your poison.

On Thu, Nov 21, 2013 at 11:48 AM, Colin Finck [e-mail] wrote:
> Hermès BÉLUSCA - MAÏTO [e-mail] wrote:
> > Ah :) This is nice !
> > By the way, I have a question: would it be useful to put such announcements
> > in some section on our website ?
>
> I wouldn't refrain from doing so if a website maintainer came to me and
> told me where to put such an announcement :)
> Colin
>
> Ros-dev mailing list
> [e-mail]
> http://www.reactos.org/mailman/listinfo/ros-dev

[Colin Finck]

Colin Finck [e-mail] wrote:
> In the process of improving the performance of our infrastructure, we
> will upgrade the first of two servers this Saturday.

Upgrade went well and all is up again :)

- Colin

[Colin Finck]

Colin Finck [e-mail] wrote:
> Upgrade went well and all is up again :)

Unfortunately, we had two unexpected crashes in the last days, which
force us to take down the server temporarily and let our ISP exchange
some hardware.

Sorry for the additional downtime..

- Colin

[Pierre Schweitzer]

So far, we are back online.

On 11/25/2013 10:30 AM, Colin Finck wrote:
> Colin Finck [e-mail] wrote:
> > Upgrade went well and all is up again :)
>
> Unfortunately, we had two unexpected crashes in the last days, which
> force us to take down the server temporarily and let our ISP exchange
> some hardware.
>
> Sorry for the additional downtime..
> Colin
>
> Ros-general mailing list
> [e-mail]
> http://www.reactos.org/mailman/listinfo/ros-general
-- 
Pierre Schweitzer [e-mail]
System Administrator
ReactOS Foundation

[Timo Kreuzer]

Not really.
Jira throws exceptions, git times out.

Am 25.11.2013 13:23, schrieb Pierre Schweitzer:
> So far, we are back online.
>
> On 11/25/2013 10:30 AM, Colin Finck wrote:
> > Colin Finck [e-mail] wrote:
> > > Upgrade went well and all is up again :)
> >
> > Unfortunately, we had two unexpected crashes in the last days, which
> > force us to take down the server temporarily and let our ISP exchange
> > some hardware.
> >
> > Sorry for the additional downtime..
> > Colin
> >
> > Ros-general mailing list
> > [e-mail]
> > http://www.reactos.org/mailman/listinfo/ros-general
>
> Ros-dev mailing list
> [e-mail]
> http://www.reactos.org/mailman/listinfo/ros-dev

[Pierre Schweitzer]

We're trying yet another time.

This time with a completely new server from Hetzner.

All services are green for the moment. There's only an obvious
bottleneck with mails.

On 11/25/2013 01:23 PM, Pierre Schweitzer wrote:
> So far, we are back online.
>
> On 11/25/2013 10:30 AM, Colin Finck wrote:
> > Colin Finck [e-mail] wrote:
> > > Upgrade went well and all is up again :)
> >
> > Unfortunately, we had two unexpected crashes in the last days, which
> > force us to take down the server temporarily and let our ISP exchange
> > some hardware.
> >
> > Sorry for the additional downtime..
> > Colin
> >
> > Ros-general mailing list
> > [e-mail]
> > http://www.reactos.org/mailman/listinfo/ros-general
>
> Ros-general mailing list
> [e-mail]
> http://www.reactos.org/mailman/listinfo/ros-general
-- 
Pierre Schweitzer [e-mail]
System Administrator
ReactOS Foundation

[Sven Barth]

On 26.11.2013 14:29, Pierre Schweitzer wrote:
> We're trying yet another time.
>
> This time with a completely new server from Hetzner.
>
> All services are green for the moment. There's only an obvious
> bottleneck with mails.

I don't know whether it's just me, but since Sunday (I last tried just 
now) I'm getting this error when I try to connect to the forum:

===

SQL ERROR [ mysqli ]

Too many connections [1040]

An sql error occurred while fetching this page. Please contact an 
administrator if this problem persists.

===

or this when I try to connect to the main page:

===

Error
The website encountered an unexpected error. Please try again later.
Error message
PDOException: SQLSTATE[08004] [1040] Too many connections in 
lock_may_be_available() (line 167 of 
/srv/www/www.reactos.org/drupal/includes/lock.inc).

===

Regards,
Sven

[Sven Barth]

Am 27.11.2013 06:42 schrieb "Sven Barth" [e-mail]:
> On 26.11.2013 14:29, Pierre Schweitzer wrote:
> > We're trying yet another time.
> >
> > This time with a completely new server from Hetzner.
> >
> > All services are green for the moment. There's only an obvious
> > bottleneck with mails.
>
> I don't know whether it's just me, but since Sunday (I last tried just

now) I'm getting this error when I try to connect to the forum:

Just in case: today it worked without problems.

Regards,
Sven