Hello everyone,

This thread will be used as a place for project admins to post updates.
If you have a comment or a question, you should post those in the appropriate threads.
There is also an Admin Updates Discussion thread for further discussion on the updates.

Thank you,
Kevin

Hello everyone,

Here is a quick update on what has been done with the server:
- Special badges have been given to Volunteers who had them
- Cancelled all of the "Ready to send" separation workunits

Currently working on:
- fixing slowness of the server
- PHP warnings on result page
- testing new binary with orbit fitting


Best,
Kevin

Hello everyone,

I will be running some server maintenance on 22 Jan 2024 at 15:00:00 UTC.
The server should only be down for 2-3 hours. I will post an update when the server is backup.

Updated on the server:
- Server should be running faster now

Working on:
- WU waiting for assimilation and validation
- Reducing the number of "Ready to send" WU
- PHP warnings on result page
- testing new binary with orbit fitting

Best,
Kevin

Hello everyone,

Update
I've been working on trying to fix some of the recent issues with the server. Here is what I have done to solve them:
- Cancelled all Separation WUs that were still creating tasks such that now the "Ready to send" tasks match (hopefully this is the last of the Separation WU's that were still creating tasks)
- Set all validated tasks that had 0 credits to "inconclusive" and set the WU to need validation so that tasks are made to validate the 0 credit task and assign them credit after validation
- Cancelled all with unsent tasks to lower the "Ready to send" number (I will be keeping the WU generator off until the numbers are lower so that this number continues to go down and tasks for validation get sent out)

I will keep watching the numbers and some of my own inconclusive tasks in order to make sure that these fixes worked. Please let me know if these fixes are working or not working for you.

Announcement
- Having done this, I will be running another server maintenance on 29 Jan 2024 at 15:00:00 UTC. Again, this should only be for a few hours at the maximum.

Question
- I do not see the PHP warnings anymore on the result pages. Is this still an issue that people are seeing? Please let me know if it is so I can take a look at it.

Working on
- finishing testing on new binary with orbit fitting
- fixing the cause for runaway task generation

Thank you,
Kevin

Hello everyone,

Update
- Removed old separation tasks for some users
- Updated the GPU requirement thread title and unpinned the GPU threads

Working on
- giving tasks needed for validation priority so credit can be given out faster
- finishing testing on new binary with orbit fitting
- fixing the cause for runaway task generation

Thank you,
Kevin

Hello everyone,

I am working on getting a new version of our code up on the server.
This new version will allow us to fit orbits on top of the other parameters we have already been fitting.
There seems to be a slight issue right now with downloading the new version and I am working on fixing the issue.

Thank you,
Kevin

Hello everyone,

I am working on getting a new version of our code up on the server.
This new version will allow us to fit orbits on top of the other parameters we have already been fitting.
There seems to be a slight issue right now with downloading the new version and I am working on fixing the issue.

Thank you,
Kevin

Update
Download error has been fixed but the error with the new version persists.
Working to get the new version working and getting new runs that have orbit fitting up on the server.

Hello everyone,

Announcement
I will be doing some server maintenance on 28 Feb 2024 at 15:00 UTC. The server should only be down for a couple of hours.
I will also be putting up new runs under a new application called Milky Way N-Body Orbit Fitting.

Update
- Fixed error causing warning messages to appear on people's profiles and on the Message Boards
- Removed the quotation mark that was appearing at the top of the home page

Working on
- GFLOPS error
- Cause of runaway WU generation

Thank you,
Kevin

Hello everyone,

Update
- We now have 3 runs in the Orbit Fitting application

Working on
- Still working on the GFLOPS issue
- Will look into getting number of CPUs used as an option in Preferences

Best,
Kevin

Hello everyone,

The person in charge of the new runs has been notified that they are erroring out and will be working to fix these ASAP.

Thanks,
Kevin

Hello everyone,

Update
- Choosing the max number of jobs and the max number of CPUs has been added to the "Preferences for this project" page
- Selecting "1" for the max number of CPUs should make the BOINC client download the single threaded version of the binary for those tasks

Please let me know if you encounter any issues with these and if the single threaded binary is not downloading for you. I was able to get the single threaded binary downloaded in testing so I am hoping there is no issues there.

Working on
- GFLOPs estimate fix is almost done and will be implemented shortly after some testing
- Still looking into the low credit issued for some tasks
- Will look into the difference in instructions between the Linux and Windows executable

Thanks,
Kevin

Hello everyone,

Update
- Wording for the new settings that were enabled in "Preferences for this project: MilkyWay@home Preferences" has been changed to hopefully be more clear
- After some testing, the single-threaded executable and multi-threaded executables running on one thread have very little difference in between them with the MT executable even being slightly faster in most cases
- MilkyWay N-Body Orbit Fitting Application executable has been changed to version 1.87, but the executable itself is exactly the same. In order to eliminate the mix of ST and MT executables that people were receiving, the new version of the executables have been set up so that all four are multi-threaded although two are labeled at single-threaded (one for Windows and one for Linux). From testing, when I set the max number of threads for each task to "1", my BOINC client crunched the tasks on one core successfully. When I set the max number of threads for each task to anything above "1", it crunched task at the max number I had set or when set to "No Limit", ran on the max number of threads for my computer.
- Default for "Max # of simultaneous MilkyWay@home tasks" is set to "10"
- Default for "Max # of threads for each MilkyWay@home task" is set to "16"

I hope this fixes all the issues, allows people to crunch WUs on one thread and makes it easier for users to change those settings.

Thank you,
Kevin

Hello everyone,

It was just announced that there will be a campus-wide network outage on 27 June 2024 from 10:00:00 to 10:30:00 UTC.
Connection to the server will be lost during that time.

Thank you,
Kevin

Hello everyone,

Update
There has been a bug in the N-body simulation code that has required immediate attention and taken up most of my time.
Once this bug is fixed, I will be able to focus again on some of the issues that were happening with the server.

Announcement
I will be performing server maintenance on 3 September 2024 at 14:00 UTC. The server should only be down for a few hours.

Thank you,
Kevin

Hello everyone,

Announcement
The server needs to be rebooted. This will happen on 6 September 2024 at 14:00 UTC. If this fixes the issue, the server will be down for less than an hour.

Thanks,
Kevin