remove the conda/mamba based updater. #2495

ReimarBauer · 2024-08-30T13:27:33Z

Show in About a new version available.

Purpose of PR?:

Partial fix of #2256

removes the updater, because it is not sufficient when we release a pinning or when we install by pixi or a different package manager

Does this PR introduce a breaking change?
removes a feature

Show in About a new version available.

…o remove_updater_i2256

matrss

Great to see this! Just some quick comments at first glance. Also, it doesn't actually address the issue described in #2256, the explicit reference to CONDA_PREFIX is still there.

mslib/utils/release_info.py

mslib/mscolab/app/__init__.py

Co-authored-by: Matthias Riße <9308656+matrss@users.noreply.github.com>

different states for the release check

ReimarBauer · 2024-09-01T06:54:00Z

Great to see this! Just some quick comments at first glance. Also, it doesn't actually address the issue described in #2256, the explicit reference to CONDA_PREFIX is still there.

I know, it is a first step, this part needs further changes. There needs more based on the mscolab_settings.

ReimarBauer · 2024-09-02T05:34:56Z

There will be a seperate PR applying changes to the gallery workflow. This needs currently an update of the docker image.
I prefer to use the way how RTFD did it.

docs/installation.rst

mslib/utils/qt.py

matrss · 2024-09-02T08:33:11Z

mslib/utils/release_info.py

+    no_new_release_found = f"{datetime.date.today()}: No new release found."
+    try:
+        # we use the @concurrent.process(timeout=1) on the function to stop requests trying to establish a connection
+        latest_release = get_latest_release().result()


Immediately calling .result() IMO defeats the purpose of using pebble at all. What does this do that isn't already accomplished by just the timeout of requests.get?

requests.get timeout is used when the connection is established.

But it retrys 10 seconds until it gives up when there is no connection possible.

This is far to long for an user clicking on about, and on the server the startup or restart can be blocked too.

Do you have an example to reproduce the issue you see with it? The requests.get timeout specifies both the time to wait for a connection to be established and the time to wait for responses from the server. With timeout=(1, 1) it should also fail after at most 1 second if there is no connection possible (well, unless the target domain has multiple IP addresses, but that doesn't seem to be the case here).

The only other situation that I am aware of where the request might take longer is when the server is responding very slowly, waiting for almost the timeout before sending another few bytes. That seems to be an unlikely edge case though, especially when accessing GitHub's API?

Just load the requests module and disconnect (or have no) internet before you execute

requests.get('https://duckduckgo.com/', timeout=(1,1))

with a connection available this shows a 200, without you wait until you get a requests.exceptions.ConnectionError

We can use MSS offline, we should not add something which can block this.

This behaves a bit different, it is faster than the about dialog

Let's look on the gifcycle how it can block my system by commenting of the decorator and starting the local server

Which network error do you have? and which requests version, I've just seen I hadn't the most recent.

Well we change our stack for our needs. When we can replace the other worker usages with it after the GSOC merges it would be usefull too.

I look if that concurrent.futures also helps.

With requests 2.32.3 pyhd8ed1ab_0 conda-forge

seems we need to uplift the pinning

it seems to work

During handling of the above exception, another exception occurred: Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/home/reimarb/Miniforge/envs/mssdev/lib/python3.11/site-packages/requests/api.py", line 73, in get return request("get", url, params=params, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/reimarb/Miniforge/envs/mssdev/lib/python3.11/site-packages/requests/api.py", line 59, in request return session.request(method=method, url=url, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/reimarb/Miniforge/envs/mssdev/lib/python3.11/site-packages/requests/sessions.py", line 589, in request resp = self.send(prep, **send_kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/reimarb/Miniforge/envs/mssdev/lib/python3.11/site-packages/requests/sessions.py", line 703, in send r = adapter.send(request, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/reimarb/Miniforge/envs/mssdev/lib/python3.11/site-packages/requests/adapters.py", line 700, in send raise ConnectionError(e, request=request) requests.exceptions.ConnectionError: HTTPSConnectionPool(host='api.github.com', port=443): Max retries exceeded with url: /repos/Open-MSS/MSS/releases/latest (Caused by NameResolutionError("<urllib3.connection.HTTPSConnection object at 0x7862fb9f88d0>: Failed to resolve 'api.github.com' ([Errno -3] Temporary failure in name resolution)"))

mamba list -r

2024-09-02 18:09:17 (rev 6) certifi {2024.7.4 (conda-forge/noarch) -> 2024.8.30 (conda-forge/noarch)} requests {2.31.0 (conda-forge/noarch) -> 2.32.3 (conda-forge/noarch)} 2024-09-02 18:16:53 (rev 7) requests {2.32.3 (conda-forge/noarch) -> 2.32.1 (conda-forge/noarch)} 2024-09-02 18:18:25 (rev 8) requests {2.32.1 (conda-forge/noarch) -> 2.31.0 (conda-forge/noarch)}

Maybe the problem was only on my system? It is also gone currently with the old version.

Which network error do you have?

As mentioned above:

requests.exceptions.ConnectionError: HTTPSConnectionPool(host='duckduckgo.com', port=443): Max retries exceeded with url: / (Caused by NameResolutionError("<urllib3.connection.HTTPSConnection object at 0x77467b735540>: Failed to resolve 'duckduckgo.com' ([Errno -2] Name or service not known)"))

So I am getting a "Errno -2 Name or service not known" while you are getting an "Errno -3 Temporary failure in name resolution".

I still can't reproduce the problem I had before. But I always now have Errno -3 independent from cable or wireless disconnection

matrss · 2024-09-02T08:34:40Z

Great to see this! Just some quick comments at first glance. Also, it doesn't actually address the issue described in #2256, the explicit reference to CONDA_PREFIX is still there.

I know, it is a first step, this part needs further changes. There needs more based on the mscolab_settings.

That's fine, but then the PR description can't mention that it fixes that issue, when it really doesn't.

ReimarBauer · 2024-09-02T11:06:14Z

It doesn't fix the name but in the discussion in the issue the problem of what needs to be changed has more facettes.

matrss · 2024-09-02T11:39:16Z

It doesn't fix the name but in the discussion in the issue the problem of what needs to be changed has more facettes.

Yes, but merging this PR would auto-close that issue, while the issue isn't solved. And that is just plain wrong. If you want an issue attached to this PR, please make removing the updater a separate issue, since that was only brought up in #2256, but not the main point of that issue, or make removing the usage of CONDA_PREFIX part of this PR.

ReimarBauer · 2024-09-02T12:25:06Z

It doesn't fix the name but in the discussion in the issue the problem of what needs to be changed has more facettes.

Yes, but merging this PR would auto-close that issue, while the issue isn't solved. And that is just plain wrong. If you want an issue attached to this PR, please make removing the updater a separate issue, since that was only brought up in #2256, but not the main point of that issue, or make removing the usage of CONDA_PREFIX part of this PR.

It is only a matter of the merge order. When it is already closed, it is connected which is also valueable.

from the issue "Some code assumes that MSS is installed with conda or mamba. This happens e.g. here:"

for me e.g. means lookup also others.

matrss · 2024-09-02T12:39:29Z

OK, makes sense, then both this PR and #2499 must be merged at the same time, as only the combination of both addresses #2256.

ReimarBauer · 2024-09-02T13:04:23Z

OK, makes sense, then both this PR and #2499 must be merged at the same time, as only the combination of both addresses #2256.

It is also common in the "unperfect" world of github to add comments to already closed issues. I splitted the PR from an undo or reading perspective.

matrss

I've updated the PR description to reflect that this is a partial fix of said issue. LGTM now.

ReimarBauer added 7 commits August 30, 2024 15:23

remove the conda/mamba based updater.

a06f12f

Show in About a new version available.

Merge branch 'develop' into remove_updater_i2256

5c19346

codespell

d7e6906

Merge branch 'remove_updater_i2256' of github.com:ReimarBauer/MSS int…

6d06aa2

…o remove_updater_i2256

docs updated

9ea7aaf

remove more updater code

d4a8ee8

flake8

e9be6c9

ReimarBauer requested review from joernu76 and matrss August 30, 2024 14:20

joernu76 approved these changes Aug 30, 2024

View reviewed changes

matrss requested changes Aug 30, 2024

View reviewed changes

mslib/utils/release_info.py Outdated Show resolved Hide resolved

mslib/utils/release_info.py Outdated Show resolved Hide resolved

mslib/mscolab/app/__init__.py Outdated Show resolved Hide resolved

ReimarBauer and others added 2 commits September 1, 2024 07:34

Update mslib/utils/release_info.py

8bdf751

Co-authored-by: Matthias Riße <9308656+matrss@users.noreply.github.com>

introduce timeout_decorator

0f129dd

different states for the release check

ReimarBauer added 3 commits September 1, 2024 12:05

use pebble

d3a10b8

improved comment

6f4991f

add pebble also to the RTFD env

953faca

matrss requested changes Sep 2, 2024

View reviewed changes

ReimarBauer added 2 commits September 2, 2024 16:01

update

0a0db83

remove pebble, not needed

d9582a4

matrss approved these changes Sep 5, 2024

View reviewed changes

ReimarBauer merged commit ee9e832 into Open-MSS:develop Sep 6, 2024
10 of 11 checks passed

ReimarBauer deleted the remove_updater_i2256 branch September 6, 2024 06:17

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

remove the conda/mamba based updater. #2495

remove the conda/mamba based updater. #2495

ReimarBauer commented Aug 30, 2024 •

edited by matrss

Loading

matrss left a comment

ReimarBauer commented Sep 1, 2024

ReimarBauer commented Sep 2, 2024

matrss Sep 2, 2024

ReimarBauer Sep 2, 2024

matrss Sep 2, 2024 •

edited

Loading

ReimarBauer Sep 2, 2024 •

edited

Loading

ReimarBauer Sep 2, 2024

ReimarBauer Sep 2, 2024

ReimarBauer Sep 2, 2024 •

edited

Loading

ReimarBauer Sep 2, 2024

matrss Sep 3, 2024

ReimarBauer Sep 4, 2024

matrss commented Sep 2, 2024 •

edited

Loading

ReimarBauer commented Sep 2, 2024

matrss commented Sep 2, 2024 •

edited

Loading

ReimarBauer commented Sep 2, 2024

matrss commented Sep 2, 2024

ReimarBauer commented Sep 2, 2024

matrss left a comment

remove the conda/mamba based updater. #2495

remove the conda/mamba based updater. #2495

Conversation

ReimarBauer commented Aug 30, 2024 • edited by matrss Loading

matrss left a comment

Choose a reason for hiding this comment

ReimarBauer commented Sep 1, 2024

ReimarBauer commented Sep 2, 2024

matrss Sep 2, 2024

Choose a reason for hiding this comment

ReimarBauer Sep 2, 2024

Choose a reason for hiding this comment

matrss Sep 2, 2024 • edited Loading

Choose a reason for hiding this comment

ReimarBauer Sep 2, 2024 • edited Loading

Choose a reason for hiding this comment

ReimarBauer Sep 2, 2024

Choose a reason for hiding this comment

ReimarBauer Sep 2, 2024

Choose a reason for hiding this comment

ReimarBauer Sep 2, 2024 • edited Loading

Choose a reason for hiding this comment

ReimarBauer Sep 2, 2024

Choose a reason for hiding this comment

matrss Sep 3, 2024

Choose a reason for hiding this comment

ReimarBauer Sep 4, 2024

Choose a reason for hiding this comment

matrss commented Sep 2, 2024 • edited Loading

ReimarBauer commented Sep 2, 2024

matrss commented Sep 2, 2024 • edited Loading

ReimarBauer commented Sep 2, 2024

matrss commented Sep 2, 2024

ReimarBauer commented Sep 2, 2024

matrss left a comment

Choose a reason for hiding this comment

ReimarBauer commented Aug 30, 2024 •

edited by matrss

Loading

matrss Sep 2, 2024 •

edited

Loading

ReimarBauer Sep 2, 2024 •

edited

Loading

ReimarBauer Sep 2, 2024 •

edited

Loading

matrss commented Sep 2, 2024 •

edited

Loading

matrss commented Sep 2, 2024 •

edited

Loading