- Welcome to eFMer - BoincTasks and TThrottle forum.
eFMer - BoincTasks and TThrottle forum
News:
Follow BoincTasks on Twitter
Facebook Visit our website here.
BoincTasks cloud login is working again
Recent posts
#91
Questions / Re: BoincTasks JS Beta 2.2 Rel...
Last post by fred - October 19, 2023, 02:17:43 AMKeep on adding, I will resolve all issues eventually.
For now I'm doing something entirely different.
For now I'm doing something entirely different.
#92
Questions / Re: TThrottle support for only...
Last post by fred - October 19, 2023, 02:15:13 AM32 is as far as I can go.
Above 32 I have to rewrite the driver and that's very expensive, probably in the thousands of dollars.
Microsoft made it much more difficult to add drivers.
Fred
Above 32 I have to rewrite the driver and that's very expensive, probably in the thousands of dollars.
Microsoft made it much more difficult to add drivers.
Fred
#93
Questions / Re: No GPU temperature shown
Last post by fred - October 19, 2023, 02:13:17 AMOK What works works. 
#94
Questions / Re: No GPU temperature shown
Last post by KLiK - October 11, 2023, 07:51:43 PMSelecting GPU-1 or GPU-2, always gives me only single option for GPU. Can't add more on v7.7.2
But Beta v7.7.8 works OK so far, all 3 temps present of all 3 GPUs.
But Beta v7.7.8 works OK so far, all 3 temps present of all 3 GPUs.
#95
Questions / Re: TThrottle support for only...
Last post by KLiK - October 11, 2023, 07:50:18 PMQuote from: fred on January 06, 2023, 02:28:48 PMUpdated to 7.74 that supports 32 temperature cores.Started to test this Beta v7.7.8, after the summer is over. (Sorry, catched this message too late.)
In Temperature sheet, there are most of the cores written (32 out of now 48) & all 3 GPUs. Previously v7.7.2 did not support GPU at all & had to run Tech Power-up to keep track of temperature on Teslas.
Will write more, as this needs to be tested on 24/7 basis.
#96
Questions / Re: BoincTasks JS Beta 2.2 Rel...
Last post by Aurum - October 11, 2023, 02:43:01 PM22. Detach a project does not work.
23. Task/Info does not show the executable or the instruction type of the WU. E.g., may be having failures with fma but not sse2 but can't see which it is.
23. Task/Info does not show the executable or the instruction type of the WU. E.g., may be having failures with fma but not sse2 but can't see which it is.
#97
Questions / Re: BoincTasks JS Beta 2.2 Rel...
Last post by Aurum - October 08, 2023, 02:46:53 PMRecieved (sic) under Tasks/Info.
Suspended at Checkpoint all tasks to do Linux updates and saw that for long running E@h O3MD1 CPU WUs many went far past the checkpoint before suspending, up to 32 minutes. I suspect that for fast running WUs like WCG SCC that checkpoint in a few minutes it may not be noticed because they actually go around the block multiple times before suspending.
Suspended at Checkpoint all tasks to do Linux updates and saw that for long running E@h O3MD1 CPU WUs many went far past the checkpoint before suspending, up to 32 minutes. I suspect that for fast running WUs like WCG SCC that checkpoint in a few minutes it may not be noticed because they actually go around the block multiple times before suspending.
#98
Questions / Re: BoincTasks JS Beta 2.2 Rel...
Last post by Aurum - September 30, 2023, 12:44:46 PM18. On the Transfer page if you click on a row and highlight it you cannot get back to having the bottom task bar display the Retry All button. If you're lucky and park the highlight on a task that gets uploaded and is removed from the list you're back. It used to sort with tasks actively being uploaded at the top but now it seems more random. It also seems less likely an UL speed will be displayed. I had a switch go on the fritz in the night and the WCG server is broken again so I had hundreds awaiting UL. One doesn't notice when things are running faster.
19. Question: The Transfer/Elapsed column displays two times separated by a dash (00:27:11 - 5:43:24), just a time followed by a dash, or just a dash. What do the two times mean? I'm guessing the left one is time actively attempting to UL and the right one is time since it entered the UL queue.
20. Some tasks are stuck on Status/State: 01022. Update, Suspend, Resume don't change it. No idea what it means but no GPU WUs for that project will run. I guess I'll have to abort them. There's message for E@H that may be a clue, or not:
"Computer: Rig-02
45131 Einstein@Home Sep 30, 2023, 05:48:08 AM Invalid global preferences supplied, please check https://einsteinathome.org/community/forum/19 on how to fix that."
Seems State 01022 may be that the task was not started by the deadline, but that can't be since the Deadline is October 14. Aborted all of the 01022 WUs that changed to State 0626 and then Aborted. After Updating E@H WUs began DLing again. Messages say the WUs finished DLing but they're stuck in Status Downloading but none begin running. I remoted into that computer and BoincManager says the same thing. All of a sudden after about 10 minutes they started running. I searched for a BOINC State Code list but found nothing.
21. When a URL appears in the Messages can it be hyperlinked so we can click on it and open it in a new tab? Now I copy it to the Clipboard and paste it in Text Editor where I can click on it and launch it.
19. Question: The Transfer/Elapsed column displays two times separated by a dash (00:27:11 - 5:43:24), just a time followed by a dash, or just a dash. What do the two times mean? I'm guessing the left one is time actively attempting to UL and the right one is time since it entered the UL queue.
20. Some tasks are stuck on Status/State: 01022. Update, Suspend, Resume don't change it. No idea what it means but no GPU WUs for that project will run. I guess I'll have to abort them. There's message for E@H that may be a clue, or not:
"Computer: Rig-02
45131 Einstein@Home Sep 30, 2023, 05:48:08 AM Invalid global preferences supplied, please check https://einsteinathome.org/community/forum/19 on how to fix that."
Seems State 01022 may be that the task was not started by the deadline, but that can't be since the Deadline is October 14. Aborted all of the 01022 WUs that changed to State 0626 and then Aborted. After Updating E@H WUs began DLing again. Messages say the WUs finished DLing but they're stuck in Status Downloading but none begin running. I remoted into that computer and BoincManager says the same thing. All of a sudden after about 10 minutes they started running. I searched for a BOINC State Code list but found nothing.
21. When a URL appears in the Messages can it be hyperlinked so we can click on it and open it in a new tab? Now I copy it to the Clipboard and paste it in Text Editor where I can click on it and launch it.
#99
Questions / BoincTasks JS Beta 2.2 Release...
Last post by Aurum - September 28, 2023, 01:36:38 PMNote: Running on Linux Mint 20.3
1. On the Projects tab highlighting a range executes the command for the top row, maybe also the second row, and occasionally none.
2. Clicking "Ready to report" doesn't do anything.
3. We used to be able to issue a command to all WUs under a collapsed folder? E.g., change "Ready to start" to Suspend. Please reinstate.
4. If a WU is tagged Suspend at Checkpoint (S@C) and it finishes before suspending the Status changes to Suspended instead of Uploading and Ready To Report. The WU does vanish.
5. No way to remove highlighting on all rows with at least one row remaining highlighted.
6. 8. In Computers/Info popup IP shown as 127.1.0.0 not its actual IP, e.g. 192.168.1.202. Looks like you're taking it from client_state <ip_addr>127.0.1.1</ip_addr>.
7. GPU not shown. client_state:
<coproc_cuda>
<count>1</count>
<name>NVIDIA GeForce RTX 2080 Ti</name>
<available_ram>11546263552.000000</available_ram>
8. Comment: Computer Info displays far more CPU flags than seem useful for BOINC, but maybe they are to someone. I see you get them from client_state <p_features>. sudo inxi -v 3 lists the relevant ones to me: avx avx2 lm nx pae sse sse2 sse3 sse4_1 sse4_2 ssse3.
9. Didn't all suspended WUs collapse into a single folder? Be nice if they would, or not in light of #3 above.
10. When a project is Suspended on the Projects tab the Tasks tab switches status from "Running" to "Waiting to run" but it should be "Suspended".
11. Occasionally WUs marked "At checkpoint" keep running and the verbose label in Status column "| Checkpoint ⏸ Don't close or leave window" is removed. I suspect that what may be happening is that if a task is marked S@C before 1:00 the command is not honored.
12. On Computers tab Password is just two asterisks. Would like to see it.
13. Computers/cpid pulls the <host_cpid> from the computer's client_state.xml file and displays it in this column. There is also <cross_project> and the one the GRC wallet uses <external_cpid>. Recommend naming the column host_cpid. I was scratching my head thinking that's not my cpid. The <host_cpid> is used to identify individual computers (hosts) in project device and stats aggregation sites, but I never go that deep in the weeds. https://boinc.berkeley.edu/forum_thread.php?id=12602
14. Projects tab used to display the number of WUs by type: CPU_WU/GPU_WU/NCI_WU. That was useful, please restore it.
. Computers tab has no way to request a refresh. It takes a long time to fill itself out and some fields seem like they may never fill in, e.g. Local machine cpid (not that I need it). One computer says cpid undefined but when I remote in <host_cpid> is defined in its client_state.xml file. Maybe File/Restart is the way to do a refresh, yes that filled out BOINC column but did not correct cpid column.
15. When Aborting multiple tasks the task name word wraps and the popup quickly grows larger than the page. I don't need to see the name of the task since they're highlighted. I just need to see the total number tasks being Aborted.
16. Recommend renamining Tasks column V_Memory to VRAM. Since there's values listed for CPU WUs I thought this was referring to virtual memory. I see you get the VRAM from the client_state. The VRAM now totals to far greater than GPU card has. I can see the actual using VRAM committed for each WU using the Linux utility NVITOP. Not sure where you get the values. No idea what the number listed for CPU WUs means.
17. When a CPU WU is suspended the CPU Utilization seems to remain fixed at its last reading. Shouldn't it go to zero?
1. On the Projects tab highlighting a range executes the command for the top row, maybe also the second row, and occasionally none.
2. Clicking "Ready to report" doesn't do anything.
3. We used to be able to issue a command to all WUs under a collapsed folder? E.g., change "Ready to start" to Suspend. Please reinstate.
4. If a WU is tagged Suspend at Checkpoint (S@C) and it finishes before suspending the Status changes to Suspended instead of Uploading and Ready To Report. The WU does vanish.
5. No way to remove highlighting on all rows with at least one row remaining highlighted.
6. 8. In Computers/Info popup IP shown as 127.1.0.0 not its actual IP, e.g. 192.168.1.202. Looks like you're taking it from client_state <ip_addr>127.0.1.1</ip_addr>.
7. GPU not shown. client_state:
<coproc_cuda>
<count>1</count>
<name>NVIDIA GeForce RTX 2080 Ti</name>
<available_ram>11546263552.000000</available_ram>
8. Comment: Computer Info displays far more CPU flags than seem useful for BOINC, but maybe they are to someone. I see you get them from client_state <p_features>. sudo inxi -v 3 lists the relevant ones to me: avx avx2 lm nx pae sse sse2 sse3 sse4_1 sse4_2 ssse3.
9. Didn't all suspended WUs collapse into a single folder? Be nice if they would, or not in light of #3 above.
10. When a project is Suspended on the Projects tab the Tasks tab switches status from "Running" to "Waiting to run" but it should be "Suspended".
11. Occasionally WUs marked "At checkpoint" keep running and the verbose label in Status column "| Checkpoint ⏸ Don't close or leave window" is removed. I suspect that what may be happening is that if a task is marked S@C before 1:00 the command is not honored.
12. On Computers tab Password is just two asterisks. Would like to see it.
13. Computers/cpid pulls the <host_cpid> from the computer's client_state.xml file and displays it in this column. There is also <cross_project> and the one the GRC wallet uses <external_cpid>. Recommend naming the column host_cpid. I was scratching my head thinking that's not my cpid. The <host_cpid> is used to identify individual computers (hosts) in project device and stats aggregation sites, but I never go that deep in the weeds. https://boinc.berkeley.edu/forum_thread.php?id=12602
14. Projects tab used to display the number of WUs by type: CPU_WU/GPU_WU/NCI_WU. That was useful, please restore it.
. Computers tab has no way to request a refresh. It takes a long time to fill itself out and some fields seem like they may never fill in, e.g. Local machine cpid (not that I need it). One computer says cpid undefined but when I remote in <host_cpid> is defined in its client_state.xml file. Maybe File/Restart is the way to do a refresh, yes that filled out BOINC column but did not correct cpid column.
15. When Aborting multiple tasks the task name word wraps and the popup quickly grows larger than the page. I don't need to see the name of the task since they're highlighted. I just need to see the total number tasks being Aborted.
16. Recommend renamining Tasks column V_Memory to VRAM. Since there's values listed for CPU WUs I thought this was referring to virtual memory. I see you get the VRAM from the client_state. The VRAM now totals to far greater than GPU card has. I can see the actual using VRAM committed for each WU using the Linux utility NVITOP. Not sure where you get the values. No idea what the number listed for CPU WUs means.
17. When a CPU WU is suspended the CPU Utilization seems to remain fixed at its last reading. Shouldn't it go to zero?
#100
Questions / Re: BoincTasks JS beta 2.1 rel...
Last post by fred - September 24, 2023, 06:34:13 AMV 2.2 Added Context menu, easier than expected and it is indeed a better way.