Posts by Conan

1) Questions and Answers : Bugs : All errors (Message 514)
Posted 18 days ago by Conan
Post:
Thanks Christian, I now have successful tasks running and finishing on my computers now.

Much appreciated for your quick response to this issue.

Conan
2) Questions and Answers : Bugs : All errors (Message 509)
Posted 18 days ago by Conan
Post:
Can you please check if you find anything unusual in your logs?
https://boinc.loda-lang.org/loda/logs.php


I can only see a few warnings about getting close to maximum memory use. And the ones that ran to 120 minutes did so without any errors that I can see.

I am sorry but I can't see much that is useful in fault finding in the logs.

I have allowed work fetch again on that computer and might catch something useful but I can't monitor for a while as it is nearly 1 in the morning and I need to go to bed.

thanks
Conan
3) Questions and Answers : Bugs : All errors (Message 506)
Posted 18 days ago by Conan
Post:
Both shift and myself are running Linux and all work units fail.
I checked another persons Windows computer and now GolfSierra has confirmed that a number fail on Windows but most seem to work.

So is there a Linux programming issue that kills all work units?

Conan
4) Questions and Answers : Bugs : All errors (Message 503)
Posted 19 days ago by Conan
Post:
App version 220722 is available which includes a fix for the hanging tasks. New work units are available!


Sorry Christian but your new application has not helped me.

Just reported 18 work units most run for 2 hours and then get Computation Errors on all work units with "195 Exit Child Failed"

Some only run for a few minutes to 1/2 hour then same failure code.

Conan

OK now up to 26 returned and 100% have failed with 195 Exit Child Failed
I will abort the rest.
Sorry but they only part work and wont complete successfully.
5) Questions and Answers : Feature requests : Badges (Message 458)
Posted 1 Jul 2022 by Conan
Post:
Thank you for the badges, nice surprise when I logged into my account.

Conan
6) Questions and Answers : Bugs : Task validation (Message 408)
Posted 19 Jun 2022 by Conan
Post:
Yes, we also observed timeouts with 220611. To mitigate the issues, we have reduced the task size to 50% in the new app version 220612. We'd appreciate if you try it out. We'll investigate all issues.

220612 is an improvement but there is more to be done.

The majority of tasks now complete (on my hosts) at 2 hours or just after and all of those validate. [Have seen other hosts that regularly do them quicker and slower.]
I have also seen tasks that ran for 3, 4 or 5 hours that also completed and validated, they also received the appropriate credit for the longer run time.

However there are some tasks that run for the full 10 hours that then complete with the message "task loda reached time limit 36000".
Some of these tasks have been marked as 'Validate error' and got no credit but others have been marked as valid and given credit, the problem with these is that the credit is only equivalent to what a 2 hour task would have received, ie a fifth of what they should have received. In view of that I have taken to aborting tasks (when I have the opportunity) that I think (hope) might not finish before the 36000 timeout is reached. The hope is that another 1, 2 or even 3 tasks might run in that time that do get the appropriate credit. The practice isn't sustainable and the issue needs to be fixed please.

There is also the problem with the 'Error while computing' message (EXIT_CHILD_FAILED) but they are more environmental and have probably been happening since the project started. I think they are mostly down to user configuration issues with their own systems but I suspect some of them may also be due to problems at the server end, possibly network issues. I rarely get them but I see others with quite a lot.


I haven't had the time limit of 36,000 second yet but I had another 3 of the time limit reached of 11,159.30 seconds,
which was the error I was receiving with the .10 versions.

Conan
7) Questions and Answers : Bugs : Task validation (Message 398)
Posted 13 Jun 2022 by Conan
Post:
Yes, we also observed timeouts with 220611. To mitigate the issues, we have reduced the task size to 50% in the new app version 220612. We'd appreciate if you try it out. We'll investigate all issues.


Thanks Christian,

Running 2206.12 now and have had 3 failures in the first 19 work units.

This one had the same issue that 2206.10 did with the exact same time out time of 11159.30 seconds, a bit like a re-badged work unit but getting the same result.

This WU had the "exit with code 195" issue and I had one other of this type.

But mostly they are running much better that earlier versions, at least so far.

Conan
8) Questions and Answers : Bugs : Are there checkpoints? (Message 397)
Posted 13 Jun 2022 by Conan
Post:
Hi Werinbert, checkpoints are written everytime the progress is updated. Similarly to your observations, we noticed issues with long-running tasks. Therefore we have reduced the taks size to 50% in 220612. We hope that this will reduce such issues. Thanks for reporting it.


I just aborted a 2206.10 task that had run for almost 19 hours and was only about 50% complete, hopefully others do much better. I am still running a 2206.10 task and then all the rest are 2206.11 tasks, 39 of them on this machine alone.


Dump them mikey,
I dumped all my 2206.10 due to excess run time limits (I think I only managed 4 successful results), skipped version 2206.11 and now have 2206.12 which have a reduced run time so don't hit the exceed limit and I have not had any failures so far with 20 odd already run.

Conan
9) Questions and Answers : Bugs : New API errors ? (Message 382)
Posted 11 Jun 2022 by Conan
Post:
New "Exceeded elapsed time limit" with 2206.10 work units, it is now 11,159.30 seconds or 3 hours 6 minutes.

Have had at least two fail with this in the last hour.

I also posted this under Task Validation thread with a link to a failed work unit.

Conan


I am aborting all 2206.10 type work units I have down loaded.

Nearly every work unit is being aborted due to the "exceeded elapsed time limit 11,159.30 seconds".

As soon as a work unit hits 3 hours and 6 minutes it gets aborted by either the Project or Boinc (I am not sure who) and all work is lost.

I will await the next version, and try again.

Conan
10) Questions and Answers : Bugs : New API errors ? (Message 381)
Posted 11 Jun 2022 by Conan
Post:
New "Exceeded elapsed time limit" with 2206.10 work units, it is now 11,159.30 seconds or 3 hours 6 minutes.

Have had at least two fail with this in the last hour.

I also posted this under Task Validation thread with a link to a failed work unit.

Conan
11) Questions and Answers : Bugs : I am not assigned to any country. (Message 380)
Posted 11 Jun 2022 by Conan
Post:
In your account, select "Other account info" there is a box there to select your country, make your choice then hit update and your not homeless anymore.

Conan
12) Questions and Answers : Bugs : Task validation (Message 379)
Posted 11 Jun 2022 by Conan
Post:
I am still getting error "reached time limit" on my 2206.10 tasks...


I just got one on this WU

Timed out at 11,159.30 seconds.

Not the 18,000 seconds I had before, just 3 hours 6 minutes instead of 4 hours 57 minutes.

So the 2206.10 tasks still have issues.

Conan

[PS] Also had a 2206.10 that started got to 0.100% and was still at 0.100% after 2 1/2 hours with 75 Days to go, so I aborted that one.
13) Questions and Answers : Bugs : Task validation (Message 375)
Posted 10 Jun 2022 by Conan
Post:
Thanks for that Christian for that, I currently have a number of the older version 2022.06, when they flush through I will test the 2022.10 version.

Thanks
Conan

[PS] Just an aside I have seen random work units run to over 8 hours, one even validated, so they don't all stop at 4 hours 57 minutes 33 seconds (18,000 seconds).
14) Questions and Answers : Feature requests : Badges (Message 364)
Posted 9 Jun 2022 by Conan
Post:
Any strong objections against badges based on number of new and updated programs per user? We could scan the programs repository for "Submitted by ..." comments to calculate that score.


None by me, these are separate badges to the overall? Or how do propose it will work.

Conan
15) Questions and Answers : Bugs : Task validation (Message 363)
Posted 9 Jun 2022 by Conan
Post:
This "reached time limit of 18,000 seconds" is becoming an issue, I have now gotten 33 validation fails due to this problem.

On this new host I downloaded 262 work units, 151 have so far completed normally, 33 have validation fails and 27 have the "exit 195 error".
So of the 211 work units so far processed 60 have failed (28%}, that is a very high failure rate don't you think?

I noticed that the work unit seems to know from just after it starts that it is going to run a long time as the estimated run time is higher than a normal work unit.
Also after reaching a certain percentage done point (anywhere from less that 1% to over 97%) the percentage no longer advances even though the work unit is using the CPU and appears to be working.

Conan
16) Questions and Answers : Bugs : Task validation (Message 361)
Posted 8 Jun 2022 by Conan
Post:
I am seeing a number of my tasks failing validation due to "Exceeded time limit of 18,000 seconds"
They all run just short of 5 hours.

I have also seen a number of my takes that fail (probably the "fails validation" tasks) that run a lot slower than most of my other tasks.
I have a 12 core 24 thread computer with 32 GB of RAM, at times all 24 threads are using 1 GB of RAM per task.

Perhaps some of my tasks are failing (the ones I have with the error code exit 195) because they need more RAM and have a hard limit on them making them stop start trying to get more RAM and so run slower than normal?

By running slower they then hit the 18,000 second time limit and fail to validate, or just fail altogether?

Conan




©2022 LODA Language