Server Status - What it means

[JacobKlein]

Hi,

I've looked at the Server Status recently, and can't make a lot of sense of it.
https://apps.foldingathome.org/serverstats

If there's a guide that explains the page, please let me know.
Otherwise, could someone please kindly consider answering these questions?

- Is a "Job" a single work unit that a slot works on?
- Is "Assign Rate" mean "# Jobs being Assigned per hour"?
- For the jobs, is there any way to see how many are currently-assigned-out, and how many are not-yet-assigned? This would be helpful especially in times of outages.
- What is an Assignment Server?
- What is a Work Server?
- What does the Errors column mean? (Tooltip says "Total With Errors/Without", but that is not helpful. Are talking about jobs that have error'd?)
- What does the Warnings column mean? (Tooltip also unhelpful)
- What does "Has CS" mean? (Tooltip also unhelpful)
- What does the "Status" column mean? (Tooltip says "Total Assign/Accept/Down", but that is not helpful.)
- Is "Space" meaning "Space available"?
- Is there any way to see not-yet-assigned for a given server? Could it be added?

If there's any way the page could be updated with some tooltips, that sure would help a newb like me out!
Can updated the column names and tooltips, be added to the backlog?

Thanks in advance,
Jacob Klein

[jonault]

If my understanding is correct, when the folding software on your PC is ready for a new work unit, it first contacts one of the Assignment Servers. The Assignment Server sends you to one of the Work Servers to download a work unit. The Work Server is chosen by matching your hardware to the needs of the project and by taking into account the overall F@H priorities (like putting Corona virus work ahead of other projects). Different research projects each have their own Work Servers; the Assignment Server is mainly there as a central point of contact coordinating everything.

And I would assume that "job" = "work unit".

As for the rest, I do not know. (Nor do I particularly care - managing the servers is someone else's job.)

[kiore]

Due to sudden demand there are some temporary issues with downloaded projects and servers are under some unexpected strain. See the announcement here: viewtopic.php?f=24&t=32424

[JacobKlein]

My questions were about making the Server Status easier to understand.
Let's make sure that replies are focused on that, please.

[bruce]

Yes, Job=WU

Status=
- Assign --> The Assignment server can direct clients to this server for it to deliver a new WU(job) to your client. It can also accept uploaded results.
- Accept --> Assigning new work is paused but server can still accept result uploads.
- Down --> (neither of the above)

What else do you not understand?

[toTOW]

- Is a "Job" a single work unit that a slot works on?
Yes, Job = Work Unit

- Is "Assign Rate" mean "# Jobs being Assigned per hour"?
Yes

- For the jobs, is there any way to see how many are currently-assigned-out, and how many are not-yet-assigned? This would be helpful especially in times of outages.
The jobs displayed are the one available for assignment. Assigned ones are not displayed.

- What is an Assignment Server?
This is the first server that the client contacts. It checks the client request parameters and assign your client to a work server.

- What is a Work Server?
This is a server that actually host the Jobs. It is where you download the WU to process and upload the results when the WU is completed.

- What does the Errors column mean? (Tooltip says "Total With Errors/Without", but that is not helpful. Are talking about jobs that have error'd?)
This column is mostly used by server maintainers. It displays what is not going well with the server : server timed out, bas configuration, ...

- What does the Warnings column mean? (Tooltip also unhelpful)
Same as errors column, but with non blocking errors.

- What does "Has CS" mean? (Tooltip also unhelpful)
CS means Collection Server. It's a backup server that will be used to return the results if the Work Server is not available.

- What does the "Status" column mean? (Tooltip says "Total Assign/Accept/Down", but that is not helpful.)
See bruce's answer.

- Is "Space" meaning "Space available"?
It's the space left on the hard drives of the server. This space is used to store WU to distribute and results returned.

- Is there any way to see not-yet-assigned for a given server? Could it be added?
This is already what is displayed as Jobs.

[JacobKlein]

Thank you for the replies!

To make the page more useful for crunchers, would you consider the following changes?

- Add a note above the Jobs grid to say "A job is a work unit. The values in the grid are jobs that are available for assignment."
- Add a tooltip to "Type" column to say "AS = Assignment Server: Assigns clients to work servers to get work. WS = Work Server: Sends/receives jobs. CS = Collection Server: Can receive jobs as a backup server."
- Add a tooltip to "Errors" column to say "Server errors. Useful for system administrators."
- Add a tooltip to "Warnings" column to say "Server warnings. Useful for system administrators."
- Add a tooltip to "Has CS" column to say "Has a Collection Server available."
- Add a tooltip to "Status" column. I don't know what this should say, because I still don't fully understand it. Is it showing numbers of other servers? I understand admins know what it means, but want a way for crunchers to know what it means too, and that it's not talking about Jobs for instance. Maybe something like "Total Assign/Accept/Down. Shows number of related servers that are: Assign (able to Assign work), Accept (able to Accept work), and Down (not online)."
- Add a tooltip to "Space" column to say "Remaining space available on the server."

Low priority for sure, but ... I'm trying to get excited about coming back to Folding@home, and this server status page is certainly not intuitive enough for a cruncher.

Please consider improving it!

Thanks,
Jacob Klein

[toTOW]

This page is being rebuilt to be integrated into a new suite of tools. We'll see how they can integrate your suggestions to the new page.

[JacobKlein]

That is excellent news. Thank you!

[JimboPalmer]

Could it be added?

The answer to 'could' is almost always yes.

Currently, F@H staff is coping with 10 to 1 growth in less than a week. I bet they are deciding to only put out large fires, not little ones.

This is not a fire. No volunteer needs to see the servers and the staff know what it all means, you just want to see this, it won't actually get more science done.

And this comes back to the Biochemists in charge priorities. If it gets more science done, they want it. If it is prettier, they don't really care.

They pay one programmer, Joseph Coffland. He writes 3 client applications (Windows, Mac, Linux) which have to stay in sync. They communicate with both assignment servers and work servers so that code which he writes must allow remain compatible. I suspect that if he had time and the biochemist had money, he wold be who writes stats code, but there is no sign either of those two have ever happened.

He is also the Web developer, and Ideally would write the Viewer program, but I bet the Biochemists never pay him to do so. Busy guy.

Meanwhile everyone here is just a volunteer, answering to the best of our abilities.

[JacobKlein]

Hi Jimbo,

I understand the staff is at critical load, and now isn't the right time to make changes you deem trivial.

However, as someone who recently re-joined F@h, I saw an opportunity for improvement, and put forth effort to learn, document, and politely request.

So, it is what it is, and I hope we see improvement. Just like all things in life

Have a good day.

[bruce]

Yes, your suggestion was made in good faith. Thank you.

Most (but not all) of FAH's software is in what is called an Open Software development repository. Experienced programmers can voluntarily submit suggested improvements which will be reviewed and, if appropriate, incorporated int a future revision of the code. I've never seen a similar arrangements for web pages, but why not? Depending on what's behind that web page, couldn't an experienced web page designer incorporate the suggested changes and then somehow submit it to the original author for eventual replacement of the existing page?

[X-Wing]

What do the different shadings of servers mean? As in, why are some grayed out and others not?

[anandhanju]

The ones with the white background are Assignment servers (AS). These are servers that the client connects to with requests for new work. The AS then routes the client request to an approrpiate Work server (WS) based on a number of factors. The status column will show the number of work servers it connects to (Assign/ Accept/ Down -- see below)

The ones in light grey are Work servers that are actively assigning work as well as accepting returned results. The status column will show "Assign" for these.
The ones in dark grey are work servers that are only accepting returned results and do not have new work to be assigned. The status column will show "Accept" for these.
The ones in super dark grey are work servers that are offline. The status column will show "Offline" for these. We currently do not see any in this status but they'll show this when the server is undergoing maintenance or is unreachable for whatever reason.

[foldinghomealone2]

Maybe just implement a link on the webpage to this thread as everything is explained in detail