I have the same strange behaviour, progress is stuck at 89.980%, while the 'Elapsed' time seem to count down (by second) from 05:05 to 05:04, and jumping back to 05:05 for the last two hours. I understand that progress can appear static by a lack of checkpoint/display frequency, but while the elapsed time does not need checkpoint information, it could be easily calculated accurately at any given moment. This looks like a bug.
Application: Gamma-ray pulsar search #5 1.08 (FGRPSSE)
Name: LATeah0035F_56.0_8_-1.8e-11
State: Running
Received: Monday, 17 September 2018 at 22:28:35
Report deadline: Monday, 01 October 2018 at 22:28:28
Estimated computation size: 105,000 GFLOPs
CPU time: 04:50:54
CPU time since checkpoint: 00:00:00
Elapsed time: 05:05:13
Estimated time remaining: 00:46:55
Fraction done: 89.980%
Virtual memory size: 4.10 GB
Process ID: 12768
Progress rate: 16.920% per hour
Executable: hsgamma_FGRP5_1.08_x86_64-apple-darwin__FGRPSSE
The apps (and data) these days are likely to be quite different to what existed more than 4 years ago so it won't be the same issue. In these situations it's always best to start a new thread.
For the current apps and data, ~90% is the point of separation between the two distinct stages of the overall calculations. The first stage does all the hard work of identifying potential candidate signals. The follow-up stage creates a 'toplist' of the 10 most likely candidates and does a more thorough analysis of them. The follow-up stage does create further checkpoints which you should be able to see by using the 'properties' button. Even though there is progress, you wont see any indication through %done until the very end when it will jump to 100%.
There is something wrong with the data you have posted and with the description you have provided. You talk about Elapsed time of 05:05 and 05:04 which looks like minutes and seconds. The properties data show the elapsed time as 05:05:13 - hrs:mins:secs as it should be. Did you leave out the secs in your description or were you really talking about the Estimated time remaining? It's very likely that the remaining time can be quite wrong during the follow-up stage. This is quite normal. The Elapsed Time should always be accurate - to the second.
The thing that seems strange is that you show an entry:-
CPU time since checkpoint: 00:00:00
Should that actually read "CPU time at last checkpoint"? If so that might mean that the task has been running for over 5 hours and has not checkpointed in all that time. Your computers are hidden (standard setting these days) so we can't check to see how powerful your computer is, in order to decide if the problem is just a very old and slow computer, or something else.
The other thing that points to a problem with your machine is the virtual memory size. You have 4.10GB I checked one of my tasks running on Linux. It was 413 MB.
What happens if you stop and restart BOINC? Have you tried that? If you haven't, you should and if it restarts right from the beginning then you really don't have any checkpoints to restart from. In that case, a full reboot of the machine might be in order before you try again. It's unlikely to be a problem with the data (or even the app) since there are no other similar reports I'm aware of.
Progress Stucks in 16.666%
)
Not all Einstein apps display frequent progress, or checkpoint frequently,
Wait, and you'll see it jump to the next checkpoint/% done.
Claggy
Thanks for your answer! You
)
Thanks for your answer! You are absolutely right!
Hi, I have the same strange
)
Hi,
I have the same strange behaviour, progress is stuck at 89.980%, while the 'Elapsed' time seem to count down (by second) from 05:05 to 05:04, and jumping back to 05:05 for the last two hours. I understand that progress can appear static by a lack of checkpoint/display frequency, but while the elapsed time does not need checkpoint information, it could be easily calculated accurately at any given moment. This looks like a bug.
Application: Gamma-ray pulsar search #5 1.08 (FGRPSSE)
Name: LATeah0035F_56.0_8_-1.8e-11
State: Running
Received: Monday, 17 September 2018 at 22:28:35
Report deadline: Monday, 01 October 2018 at 22:28:28
Estimated computation size: 105,000 GFLOPs
CPU time: 04:50:54
CPU time since checkpoint: 00:00:00
Elapsed time: 05:05:13
Estimated time remaining: 00:46:55
Fraction done: 89.980%
Virtual memory size: 4.10 GB
Process ID: 12768
Progress rate: 16.920% per hour
Executable: hsgamma_FGRP5_1.08_x86_64-apple-darwin__FGRPSSE
Welcome to
)
Welcome to Einstein@Home.
The apps (and data) these days are likely to be quite different to what existed more than 4 years ago so it won't be the same issue. In these situations it's always best to start a new thread.
For the current apps and data, ~90% is the point of separation between the two distinct stages of the overall calculations. The first stage does all the hard work of identifying potential candidate signals. The follow-up stage creates a 'toplist' of the 10 most likely candidates and does a more thorough analysis of them. The follow-up stage does create further checkpoints which you should be able to see by using the 'properties' button. Even though there is progress, you wont see any indication through %done until the very end when it will jump to 100%.
There is something wrong with the data you have posted and with the description you have provided. You talk about Elapsed time of 05:05 and 05:04 which looks like minutes and seconds. The properties data show the elapsed time as 05:05:13 - hrs:mins:secs as it should be. Did you leave out the secs in your description or were you really talking about the Estimated time remaining? It's very likely that the remaining time can be quite wrong during the follow-up stage. This is quite normal. The Elapsed Time should always be accurate - to the second.
The thing that seems strange is that you show an entry:-
CPU time since checkpoint: 00:00:00Should that actually read "CPU time at last checkpoint"? If so that might mean that the task has been running for over 5 hours and has not checkpointed in all that time. Your computers are hidden (standard setting these days) so we can't check to see how powerful your computer is, in order to decide if the problem is just a very old and slow computer, or something else.
The other thing that points to a problem with your machine is the virtual memory size. You have 4.10GB I checked one of my tasks running on Linux. It was 413 MB.
What happens if you stop and restart BOINC? Have you tried that? If you haven't, you should and if it restarts right from the beginning then you really don't have any checkpoints to restart from. In that case, a full reboot of the machine might be in order before you try again. It's unlikely to be a problem with the data (or even the app) since there are no other similar reports I'm aware of.
Cheers,
Gary.