Apprendo con una certa felicità che Bryan, il co-amministratore di Riesel Sieve, ha comunicato sul forum di backup del progetto che è riuscito ad avere un contatto diretto con Lee, l'amministratore che era stato dato per disperso Mentre era intento a mettere in piedi una sorta di progetto di backup (cioè ritornare al vecchio progetto LLRnet che usa una struttura più semplice), Bryan è stato contattato da Lee che è finalmente riuscito a tornare online.
Dopo averci parlato, è riuscito (finalmente) a postare un update (il 1° Agosto) che ci informa su quale sarà il futuro di Riesel Sieve:
| Just an update, Lee has the servers back in his possession, although he does not have enough/long enough cables to bring them back online so I can begin the process of restoration. The servers are essentially (from what he can tell) in tact, although the router/reverse proxy's CF card is damaged & must be replaced. The goal is to begin DB work tomorrow evening (Saturday), and move on from there. More than likely, we will be reformatting/installing every server, which will add a little time to our restore schedule. Bryan |
Abbiamo inoltre altre informazioni che indicano lo stato dei lavori per il ripristino della funzionalità del progetto:
| So, this is going to sound really bizarre, and possibly contrived, but we have heard from B2!!!! Lee returned to online status at just after midnight, August 1st. He will have a statement about the outage/shutdown this weekend. In the mean time, I am preparing a project resumption plan (basically scrapping the plans I outlined Thursday). What I have so far: 1) Restore DB server, check data integrity 2) Restart web server for forum/website 3) Restart LLRNet server 4) Upgrade BOINC server to 'traditional'/'Berkley' BOINC software - since we're already in an outage, a few more days won't hurt any more 5) Restart BOINC server Ideally, 1-3 should be done by Saturday night, hopefully. As we have more details, I will post them. |
Infine, una risposta a chi gli chiede cosa succederà a tutte le WU elaborate successivamente al 28 Giugno ed oramai scadute:
| I understand your concern - I unfortunately can't give an answer one way or the other on that specific question, though. It largely depends upon what state the data is in when I can get access to it. In theory, the server should resume exactly where it was on June 28th. There is a possibility that workunits will have been marked 'too late' and regenerated, if the server was turned on since June 28th, but there won't be a crunched workunit to CANCEL them out. Therefore, my goal would be to find a way to credit work that was out on June 28th. That said, if the DB does not have those workunits anymore due to crash/changes/whatever, then I would have no way to credit them. I guess the simple answer is - if there's a way to, yes, we'll find a way to credit those workunits that were sent to /dev/null by BOINC or by the users. However, there's too much uncertainty to make that a definite 'yes'. Bryan |
Bryan promette che faranno tutto il possibile per garantire i dovuti crediti a chi ha elaborato le WU scadute.
Quindi, con un po' di fortuna, la prossima news su Riesel Sieve sarà per informare tutti del ripristino del progetto. Incrociamo le dita.