𝕏

Posts by jcastro

21) Message boards : Number crunching : Very slow site (Message 1104)
Posted 10 Oct 2016 by Profile jcastro
Post:
Hi!

We have solved the issue with the wordpress login page.

Regards, Joel.
22) Message boards : Number crunching : Got some beta 1.06 work! (Message 1097)
Posted 1 Oct 2016 by Profile jcastro
Post:
We are starting with small batches, now the information that the "system" give to us is sufficient. It that last upgrade, we have change the checkpoint so if you want to help us a lot, you can do a lot of "stop/pause" with the WUs of Beta. We will launch more WUs in few days.

Regards, Joel.
23) Message boards : Number crunching : Very slow site (Message 1096)
Posted 1 Oct 2016 by Profile jcastro
Post:
Hi!
Yes, we have had issues this last week. But It have to be solved now.

Best regards, Joel.
24) Message boards : Number crunching : Got some beta 1.06 work! (Message 1092)
Posted 29 Sep 2016 by Profile jcastro
Post:
If your task name starts with "BETA_28091846_3000" you can abort it.
25) Message boards : Number crunching : Got some beta 1.06 work! (Message 1090)
Posted 29 Sep 2016 by Profile jcastro
Post:
first 50 WUs had an error on the configuration file. We have cancel them and no more WUs will be sent with that configuration. The second group ( from 50 to 100) are corrects and seem to work (and end approximately in 3 hours)
26) Message boards : Number crunching : Got some beta 1.06 work! (Message 1087)
Posted 29 Sep 2016 by Profile jcastro
Post:
Yes,
We are testing a double file checkpoint. But in the other hand, we are experimenting issues with the web server.

Regards, Joel.
27) Message boards : Number crunching : No WU's available? (Message 1084)
Posted 22 Sep 2016 by Profile jcastro
Post:
We hope that in few weeks it will be work for everyone :)
28) Message boards : News : New WUs Generation will be paused (Message 1083)
Posted 22 Sep 2016 by Profile jcastro
Post:
Hi!

Sorry on the delay of the answer. Most part of our staff is teaching inside the San Jorge's University, so those day have been the start of the new course and a lot of work have maintain us away from DENIS.

Also we are doing some work in the "shadows" that will be visible soon. We are generating a new version with double check-pointing and in the future more analysis about the checkpoint file ( error failures...).

Also Jesús Carro is developing a new model which will available soon to crunch on it. This new model would imply a lot of work. Also we are regarding the best way to obtain reasonable results with less CPU resources, this is why we are delayed the start of CRN subproject.

The double checkpoint version would be available on BETA in more or less 2 or 3 days even less. And if it goes well it system will be replied to stable versions of DENIS.

Best regards, Joel.
29) Message boards : Number crunching : Android app ? (Message 1072)
Posted 26 Aug 2016 by Profile jcastro
Post:
The "problem" is not just the power. Is also the requirements to start crunching in a device ( ej over 90% battery and plugged) This could cause that some devices will compute just few minutes.

Also, our priority is a stable version for computers and later the android and ARM(Raspberry pi ) versions.

Best regards, Joel.
30) Message boards : Number crunching : DENIS_BETA v1.05 checkpoints are SOMETIMES working (Message 1067)
Posted 19 Aug 2016 by Profile jcastro
Post:
Hi!
We will add the second checkpoint for sure in a new version. If you want, you can finish the task, if you want the credit. I think that the info about the error will continue in the stderr when it finish completely.

Best regards, Joel.
31) Message boards : Number crunching : All new! (Message 1051)
Posted 7 Aug 2016 by Profile jcastro
Post:
CRLP2011EPI 1.09 Based on Beta 1.04
Beta 1.05 with more data saved on checkpoints.
CRN - Based on Beta 1.04 with new model "in development (You will get more info soon)

Best regards, Joel
32) Message boards : Number crunching : DENIS_BETA v1.04 checkpoints are working (Message 1049)
Posted 5 Aug 2016 by Profile jcastro
Post:
Hi!

We think that is stable enough. We have upload the code few days ago to github in order to permit sesef and others to optimize it.

The length of the WUs have a reason, smaller WUs doesn't give us as much errors as longest, so we have had to try with that longest ones.

In a future we will start beta testing with shorter simulations and few of them a bit longer.

We are going to launch a bunch of work in stable version of the app, so stay tuned :)

Best regards, Joel.
33) Message boards : Number crunching : DENIS_BETA v1.04 checkpoints are working (Message 1046)
Posted 3 Aug 2016 by Profile jcastro
Post:
Hi!

We have deleted the un-sent WUS and remains crunching those that are in progress. We hope that those who wants credits re-assign their host to the beta now that is stable.

We want to keep crunching the rest of the WUs in order to find if there are more bugs. Also we are working on the bagdes for BetaTesters.

As said in other post, we will send more work when we want to test new things on the app, maybe in groups of task smaller and with larger delay time to answer.

Best regards, Joel
34) Message boards : Number crunching : DENIS_BETA v1.03 & v1.04 tasks finished (pending validation) and no credit given? (Message 1042)
Posted 1 Aug 2016 by Profile jcastro
Post:
Hello!

There was a misconfiguration on the validator. It have been solved now, and it seems to work ok.

Thanks for the info!,best regards, Joel.
35) Message boards : Number crunching : DENIS_BETA v1.03 & v1.04 tasks finished (pending validation) and no credit given? (Message 1035)
Posted 30 Jul 2016 by Profile jcastro
Post:
Hi!

The number of crunchers that test the beta app (less than the "stable" version) in addition to the time cost of solve one task of the WU means that to solve a complete WU it could cost a bit more ( maybe 3 days or more depending on the distribution of work). I think that in few days another user will pick the tasks of your WUs and finish it.

In order to answer the thing of non-completed simulations in Beta, we have decided that give credit for cpu-time could be easily hackeable and could create some problems. So we are designing badges for the people that crunch in beta. WE have to decided and see how to do it, but as you say, we also think that Beta crunchers needs acknowledgement.

Best regards, Joel.
36) Message boards : Number crunching : DENIS_BETA v1.04 checkpoints are working (Message 1030)
Posted 29 Jul 2016 by Profile jcastro
Post:
Yes! We hope everything is solved, but we have to wait until the results come back to the server to determine that.

Best regards, Joel.
37) Message boards : Number crunching : Very long wus (Message 1028)
Posted 29 Jul 2016 by Profile jcastro
Post:
Hi!

We are thinking different ways to ensure checkpoints. As you have said, a double checkpoint file as a backup and check the size of the file is one of the ways that are on the table.

Writing into disk is one of the most "expensive" things in the execution of the simulation, this is why at the beginning we have tried to simplify this part. But with those ultra-long WUs it could be necessary.

Best regards, Joel.
38) Message boards : Number crunching : Completed, can't validate (Message 1020)
Posted 28 Jul 2016 by Profile jcastro
Post:
Hi!

We have canceled all the WUs that could be buggy in order to avoid to your host keep more time computing in an endless loop bug. Sorry for the inconvenience.

Best regards, Joel.
39) Message boards : Number crunching : Very long wus (Message 999)
Posted 26 Jul 2016 by Profile jcastro
Post:
I used SysInternals Process Monitor to check what the beta application does in the slot directory during a checkpoint. The sequence is as follows:

  1. File "denis_state" is not found.
  2. Creates the file "temp" with "Desired Access: Generic Write, Read Attributes" and "ShareMode: Read, Write". Successful with "OpenResult: Overwritten". Does this twice.
  3. Writes the "CP" line to "stderr.txt".
  4. Writes 1280 bytes to "temp" then successfully closes it once.
  5. Attempts to create "temp" with "Desired Access: Read Attributes, Delete, Synchronize", "Disposition: Open", "ShareMode: Read, Write, Delete". This causes a sharing violation (repeated on 5 further attempts) because the file was opened twice with "ShareMode: Read, Write", only closed once and the new open attempt is trying to add "Delete" to "ShareMode".


The obvious solution to this is to remove the spurious open of "temp".


Thanks for the info, we are trying different strategies to checkpoint to avoid the problems we have. Maybe this solution is no the correct. We are modifing a bit more.

Best regards, Joel.
40) Message boards : Number crunching : Very long wus (Message 982)
Posted 23 Jul 2016 by Profile jcastro
Post:
We have stop new simulations on CRLP2011, and we are going to do some work on the BETA.


The beta is public? Can we help you??


To help us with the beta, you have to mark "Run test applications?" in "DENIS@Home Preferences" in your profile.

We use this separate app to launch few simulations with strange parameters or new features that we want to test before add it into the "stable/official" DENIS simulator.

Best regards, Joel.


Previous 20 · Next 20