Quote:
Originally Posted by 7777  Hello experience NSS Users!
I run 48-70 Clients PC and i do mega chunk(each time 10 chunks) Server submit to Clients,so if 1 of my Clients PC failed in the midle of chunk(job),but the SP code is on that chunks from PC failed(for example) what will Server do?
Will i need to fix that PC ASAP(if i know) and do restore that chunks or the Server will regiving that chunks from the PC failed to other Clients PC?If the Server not regiving that chunks from PC failed to the other,then we lost the electronic bill for the other PC running for nothing
Thats very importand info i think all NSSPro need to know too.
Thanks and have a nice weekend
Sorry Chris! due the time zone from me and you different i can't waitting you from ICQ,when you available then my time is 2-3AM,so i'm asking these very important info i need to know here
You have a nice weekend |
Hi,
Interesing query. In theory, the scenario is the end of the world. The way
the system is coded, you have a chance to recover from the error. So with
70 computers crunch very fast. By the time you found out that there is a
computer crash on one of the nodes - the job would be 100 % error, and
you would have 5-6 jobs done, and maybe one active. All you have to do
is go and restart the PC, it will start the same chunk from the start, find
the code, submit to the server (regardless that your job is no longer active)
and the status will change from error to done.
In practice though, even one computer fails - statistically the chances to
spoil your job are very low, the same low as every second phone you do
is found at 0.1 %.
Hope this helps.
BR