Smalltalk › Pharo › Pharo Smalltalk Users

stfx.eu down

_‹ Previous Topic Next Topic _›

Classic

List

Threaded

5 messages Options

Sven Van Caekenberghe-2

stfx.eu down

Hi,

This evening I somehow broke my stfx.eu server while doing an ubuntu dist upgrade (an old AWS EC2 machine).

This means several websites / web services hosted by that machine are down, most notably mc.stfx.eu (hosting several of my packages) and ws.stfx.eu (used for shared smalltalk workspaces). A couple of Zinc tests will also fail.

For source code access, all my public repos exist both in an up to date form on GitHub (http://github.com/svenvc) as well in the original form in SmalltalkHub (http://smalltalkhub.com/#!/~SvenVanCaekenberghe) or SS3.

The main problem there is that the distributed ConfigurationsOf and Catalog entries will still point to the original. This is not a problem with git's BaselineOf definitions.

Although the machine will no longer boot, I managed to mount the original volume on another instance so I can still access the whole filesystem. This will make it possible to eventually recover most services. But this will take time as these things never happen when it is convenient. It will give me an opportunity to modernise the setup though, which is good.

Sorry for the inconvenience.

Sven

Paul DeBruicker

Re: stfx.eu down

I wonder if there is a way to get all the relevant configs in an image to
point to the smalltalkhub repo instead of the mc.stfx.eu repo temporarily.

In Pharo 6.1 if a person wants to switch away from the mc.stfx.eu domain in
their ConfigurationOf* and BaselineOf* specs you can right click on a
package in Nautilus, choose "Rewrite code" then choose "Rewrite code" again
and paste each of the below into the box and save it to rewrite the
repository definition.

RBParseTreeRewriter new
replace: '''http://mc.stfx.eu/ZincHTTPComponents''' with:
'''http://smalltalkhub.com/mc/SvenVanCaekenberghe/ZincHTTPComponents/main''';
yourself

RBParseTreeRewriter new
replace: '''http://mc.stfx.eu/Zodiac''' with:
'''http://smalltalkhub.com/mc/SvenVanCaekenberghe/Zodiac/main''';
yourself

RBParseTreeRewriter new
replace: '''http://mc.stfx.eu/Neo''' with:
'''http://smalltalkhub.com/mc/SvenVanCaekenberghe/Neo/main''';
yourself

to "fix" Zinc, Zodiac, and Neo respectively.

Sven Van Caekenberghe-2 wrote

> Hi,
>
> This evening I somehow broke my stfx.eu server while doing an ubuntu dist
> upgrade (an old AWS EC2 machine).
>
> This means several websites / web services hosted by that machine are
> down, most notably mc.stfx.eu (hosting several of my packages) and
> ws.stfx.eu (used for shared smalltalk workspaces). A couple of Zinc tests
> will also fail.
>
> For source code access, all my public repos exist both in an up to date
> form on GitHub (http://github.com/svenvc) as well in the original form in
> SmalltalkHub (http://smalltalkhub.com/#!/~SvenVanCaekenberghe) or SS3.
>
> The main problem there is that the distributed ConfigurationsOf and
> Catalog entries will still point to the original. This is not a problem
> with git's BaselineOf definitions.
>
> Although the machine will no longer boot, I managed to mount the original
> volume on another instance so I can still access the whole filesystem.
> This will make it possible to eventually recover most services. But this
> will take time as these things never happen when it is convenient. It will
> give me an opportunity to modernise the setup though, which is good.
>
> Sorry for the inconvenience.
>
> Sven

--
Sent from: http://forum.world.st/Pharo-Smalltalk-Users-f1310670.html

Offray Vladimir Luna Cárdenas-2

Re: stfx.eu down

In reply to this post by Sven Van Caekenberghe-2

Hi Sven,

Thanks for the heads up. stfx has been really useful when sharing
playgrounds in workshops and hackathons, and of course we can wait until
modernization ends.

Cheers,

Offray

On 10/12/18 16:10, Sven Van Caekenberghe wrote:

> Hi,
>
> This evening I somehow broke my stfx.eu server while doing an ubuntu dist upgrade (an old AWS EC2 machine).
>
> This means several websites / web services hosted by that machine are down, most notably mc.stfx.eu (hosting several of my packages) and ws.stfx.eu (used for shared smalltalk workspaces). A couple of Zinc tests will also fail.
>
> For source code access, all my public repos exist both in an up to date form on GitHub (http://github.com/svenvc) as well in the original form in SmalltalkHub (http://smalltalkhub.com/#!/~SvenVanCaekenberghe) or SS3.
>
> The main problem there is that the distributed ConfigurationsOf and Catalog entries will still point to the original. This is not a problem with git's BaselineOf definitions.
>
> Although the machine will no longer boot, I managed to mount the original volume on another instance so I can still access the whole filesystem. This will make it possible to eventually recover most services. But this will take time as these things never happen when it is convenient. It will give me an opportunity to modernise the setup though, which is good.
>
> Sorry for the inconvenience.
>
> Sven
>
>
>
>
>

Sven Van Caekenberghe-2

Re: stfx.eu down

In reply to this post by Sven Van Caekenberghe-2

> On 10 Dec 2018, at 22:10, Sven Van Caekenberghe <[hidden email]> wrote:
>
> Hi,
>
> This evening I somehow broke my stfx.eu server while doing an ubuntu dist upgrade (an old AWS EC2 machine).
>
> This means several websites / web services hosted by that machine are down, most notably mc.stfx.eu (hosting several of my packages) and ws.stfx.eu (used for shared smalltalk workspaces). A couple of Zinc tests will also fail.

I recovered parts of the functionality: ws.stfx.eu (shared smalltalk workspaces) and zn.stfx.eu are back, so that normal Zinc unit tests should pass everywhere.

At least all this is now running on Ubuntu 18.04.1 LTS on Pharo 7 64-bit, as it should.

Next up is http://mc.stfx.eu for my Monticello repository, but that might take more time.

> For source code access, all my public repos exist both in an up to date form on GitHub (http://github.com/svenvc) as well in the original form in SmalltalkHub (http://smalltalkhub.com/#!/~SvenVanCaekenberghe) or SS3.
>
> The main problem there is that the distributed ConfigurationsOf and Catalog entries will still point to the original. This is not a problem with git's BaselineOf definitions.
>
> Although the machine will no longer boot, I managed to mount the original volume on another instance so I can still access the whole filesystem. This will make it possible to eventually recover most services. But this will take time as these things never happen when it is convenient. It will give me an opportunity to modernise the setup though, which is good.
>
> Sorry for the inconvenience.
>
> Sven
>
>
>

Sven Van Caekenberghe-2

Re: stfx.eu down

> On 11 Dec 2018, at 21:23, Sven Van Caekenberghe <[hidden email]> wrote:
>
>> On 10 Dec 2018, at 22:10, Sven Van Caekenberghe <[hidden email]> wrote:
>>
>> Hi,
>>
>> This evening I somehow broke my stfx.eu server while doing an ubuntu dist upgrade (an old AWS EC2 machine).
>>
>> This means several websites / web services hosted by that machine are down, most notably mc.stfx.eu (hosting several of my packages) and ws.stfx.eu (used for shared smalltalk workspaces). A couple of Zinc tests will also fail.
>
> I recovered parts of the functionality: ws.stfx.eu (shared smalltalk workspaces) and zn.stfx.eu are back, so that normal Zinc unit tests should pass everywhere.
>
> At least all this is now running on Ubuntu 18.04.1 LTS on Pharo 7 64-bit, as it should.
>
> Next up is http://mc.stfx.eu for my Monticello repository, but that might take more time.

At last, http://mc.stfx.eu is back up.

I still recommend everyone to use the http://github.com/svenvc versions of my projects, as this is the better option going forward. I also intend to use these as primary repositories.

>> For source code access, all my public repos exist both in an up to date form on GitHub (http://github.com/svenvc) as well in the original form in SmalltalkHub (http://smalltalkhub.com/#!/~SvenVanCaekenberghe) or SS3.
>>
>> The main problem there is that the distributed ConfigurationsOf and Catalog entries will still point to the original. This is not a problem with git's BaselineOf definitions.
>>
>> Although the machine will no longer boot, I managed to mount the original volume on another instance so I can still access the whole filesystem. This will make it possible to eventually recover most services. But this will take time as these things never happen when it is convenient. It will give me an opportunity to modernise the setup though, which is good.
>>
>> Sorry for the inconvenience.
>>
>> Sven
>>
>