Carl\'s hard work has got my models caught up in regards to trickles, Thanks mate, I was getting a bit worried.

[mutter mode]
Mongrel hackers
[/mutter mode]

However after looking at my two coupled models, the graphs have holes in them for the delayed trickles. Credit however has been given. The other Sulpher model has all data.

My question is this.
Will this gap affect the result data?
Will this data be recovered from the 10year upload?
Should I worry about this?

---

With slab and sulphur, the intermediate trickles just said: \"I\'m still running, and this is where I\'m up to\". The data was in the end of phase uploads, all uploaded on completion of the model.

They also worked differently; if you missed a few trickles, when the next trickle IS received, the server software says: \"This is trickle No. nn, which means that there should be xx credits issued so far for this model. There are less than this, so this model gets THIS amount of credits right now.\"
The only problem was if an end of phase upload went missing; then the data for that phase would be permanently lost.

The new TCMs are different; EACH trickle contains data, hence the gaps if some go missing. But credits will be continuously calculated, the same way as with slab and sulphur.

With slab and sulphur, the intermediate trickles just said: \"I\'m still running, and this is where I\'m up to\". The data was in the end of phase uploads, all uploaded on completion of the model.

They also worked differently; if you missed a few trickles, when the next trickle IS received, the server software says: \"This is trickle No. nn, which means that there should be xx credits issued so far for this model. There are less than this, so this model gets THIS amount of credits right now.\"
The only problem was if an end of phase upload went missing; then the data for that phase would be permanently lost.

The new TCMs are different; EACH trickle contains data, hence the gaps if some go missing. But credits will be continuously calculated, the same way as with slab and sulphur.

Thanks Les.
So the gaps are there to stay. Seams strange cause to trickle\'s on the TCM were done seperatly. I would have though the data would still be there, But on consideration, All those trickles for everyone, Thats a LOT of data.
Oh well
Say La Vie.

They may still show up. Perhaps they got on the wrong plane, and are now in Rio, having the time of their lives. :)

Seriously, the trickles may be at Oxford, but the graph program may not have got to them yet.
It\'s all rather complicated, with several servers, and several programs for different things.

Les, thanks very much for the explanations contrasting coupled-model trickles with sulphur; it\'s very helpful. And of course thanks to Carl for the long night of database repair. (May the trickle-hacker fry their equipment so that they can\'t use it for harassment any more.)

are the \"trickle gaps\" filled in? I think everything is caught up. I redid the trickle server-side processing to be a \"daemon\" and keep up to date. I also save all the trickle zip files so in theory I could rebuild the trickle database (it takes forever though!)

Looking at the graphs against his coupled models, they look OK.

---

I'm a volunteer and my views are my own.
News and Announcements and FAQ