climateprediction.net home page
Posts by Milo Thurston

Posts by Milo Thurston

61) Message boards : Number crunching : No more regional models? (Message 40683)
Posted 16 Sep 2010 by Profile Milo Thurston
Post:
Apparently a small number more are required.
62) Message boards : Number crunching : No more regional models? (Message 40681)
Posted 16 Sep 2010 by Profile Milo Thurston
Post:
I have asked the physicists whether they need any more of these models issuing right now and I am waiting for their reply. They may ask me to delay for a bit in order to get more FAMOUS models run.
63) Message boards : Number crunching : Another Upload Problem (Message 40679)
Posted 16 Sep 2010 by Profile Milo Thurston
Post:
I have been able to ssh into the Oregon server. The file_upload_handler is clearly working as files are coming in and the upload logs agree. The result templates have the correct URLs for the file_upload_handler on this system. Therefore, I can't find any obvious problem for me to fix.

Are those of you having problems able to see a response from this URL if you check in a browser?


http://boinc1.coas.oregonstate.edu/cpdn_cgi_main/file_upload_handler

You should see some XML like this:

<data_server_reply>
    <status>1</status>
    <message>no command</message>
</data_server_reply>
64) Message boards : Number crunching : Computer wasting multiple models (Message 40674)
Posted 15 Sep 2010 by Profile Milo Thurston
Post:
The most recent ones are done - apologies for the delay.
65) Message boards : Number crunching : Upload problem (Message 40663)
Posted 13 Sep 2010 by Profile Milo Thurston
Post:
I've now put a small NAS unit in the server room where climateapps1 is stored and I'm slowly copying data to it. This is not an ideal solution but it is small and cheap so I was actually able to get hold of it in a matter of days rather than months.

Hopefully the server can be re-started later today.
66) Message boards : Number crunching : Upload problem (Message 40642)
Posted 9 Sep 2010 by Profile Milo Thurston
Post:

Good, learned something new - remote servers (outside Oxford) may cause hold-ups (as servers do) but not reported on the status page.


I've added that one to the status page, although I can't guarantee that the result will always be accurate.
67) Message boards : Cafe CPDN : welcome to join our new team: Crises Killer (Message 40614)
Posted 7 Sep 2010 by Profile Milo Thurston
Post:
Actually, seems this forum do not have this board to post...


Here it is:
http://climateapps2.oucs.ox.ac.uk/cpdnboinc/forum_thread.php?id=5391#27593
68) Message boards : Number crunching : Credit anomalies (Message 40583)
Posted 3 Sep 2010 by Profile Milo Thurston
Post:
Database reloading complete - I'm turning the daemons back on again.
Fingers crossed!
69) Message boards : Number crunching : Credit anomalies (Message 40582)
Posted 3 Sep 2010 by Profile Milo Thurston
Post:
I've also lost more than 60.000 Points!!!

Will it help you, if we stop all activity so you must not wait 5 minutes for a query result?


Don't worry about it - the problem is only because MySQL can be rather slow and we've got some pretty big tables.
I've left the connection to the database running so that users can still access this board, but there shouldn't be any BOINC clients connecting.
70) Message boards : Number crunching : Credit anomalies (Message 40579)
Posted 3 Sep 2010 by Profile Milo Thurston
Post:
Maybe it's just too early...


Unfortunately so. I've been having very great difficulty with the backups, I'm afraid. I've managed to restore some data, but there's still quite a bit more to do.

Even a simple query, e.g. to count the number of entries in a table, takes a little over 5 minutes and a restore attempt or dump several hours.
71) Message boards : Number crunching : Credit anomalies (Message 40573)
Posted 3 Sep 2010 by Profile Milo Thurston
Post:

Thanks Milo, For your work on this. As an ex-programmer I know what you are up against. Keep up the good work.


Thanks!
Some data are restored but there's a bit of work to do yet.
72) Message boards : Number crunching : Credit anomalies (Message 40564)
Posted 2 Sep 2010 by Profile Milo Thurston
Post:
A restore is still running - hopefully it will have finished by tomorrow morning and I'll be able to get to work trying to fix the mess.

Apologies again for the huge delays trying to get this data restored.
73) Message boards : Number crunching : Credit anomalies (Message 40555)
Posted 2 Sep 2010 by Profile Milo Thurston
Post:
Milo, you said: 'Also, the vanished results should be old ones which have already finished.'

But if MacDitch is right about the task and WU numbers for his currently crunching model, it must mean that in at least this case the task page for an uncompleted model has disappeared.


Indeed they should be…
I will know more when a backup has been successfully restored.
74) Message boards : Number crunching : Credit anomalies (Message 40551)
Posted 2 Sep 2010 by Profile Milo Thurston
Post:
Silly question, but what will happen when I try to trickle-up models that have dissappeared? I'm assuming that I shouldn't try this until the restore has happened...

My current model is:
Task ID: 7934758
Work Unit ID: 6212331
Computer: 865981


You shouldn't be able to, as I turned the BOINC daemons off once it was clear there was a serious problem. I'll turn them back on again when I've dealt with it.

Also, the vanished results should be old ones which have already finished.
75) Message boards : Number crunching : Credit anomalies (Message 40518)
Posted 1 Sep 2010 by Profile Milo Thurston
Post:
It's probably going to take another 24 hours or more to sort everything out. I'm restoring the last backup overnight so that we can identify what's gone missing.
76) Message boards : Number crunching : Credit anomalies (Message 40475)
Posted 1 Sep 2010 by Profile Milo Thurston
Post:
Have a look at the "Total Credit, last months" graphs of myself and Doc.Brown.

Although the totals are similar, Doc.Brown is a relatively new contributor and seems unaffected by the loss of credit.


Thanks - that would fit the hypothesis that it's a problem with missing results in the archive, which is the problem we've seen before.
77) Message boards : Number crunching : Credit anomalies (Message 40473)
Posted 1 Sep 2010 by Profile Milo Thurston
Post:
I'll be restoring some data from backups as soon as I've copied it over. This should allow me to determine what's gone missing and to fix as necessary.
78) Message boards : Number crunching : Credit anomalies (Message 40468)
Posted 1 Sep 2010 by Profile Milo Thurston
Post:
OK, it looks as if there is clearly a problem.

We had credit problems around this time last year when we had to do a database archive. I'm investigating now to see if the same thing has occurred. If so, then there will be further outages over the next day or so whilst I deal with it.

Apologies for the inconvenience.
79) Message boards : Number crunching : Credit anomalies (Message 40463)
Posted 1 Sep 2010 by Profile Milo Thurston
Post:
EDIT:

I'm working on a fix for this problem as my top priority. There's enough information now so you don't need to reply to this thread. Credits should return to normal once the fix is complete, which I expect to take at least 24 hours.

Thanks for your patience!
80) Message boards : Number crunching : Computer wasting multiple models (Message 40460)
Posted 31 Aug 2010 by Profile Milo Thurston
Post:

All the others in Ageless's two posts need the email and to be minussed though.


Done.


Previous 20 · Next 20

©2024 climateprediction.net