Pages: [1] :: one page |
|
Author |
Thread Statistics | Show CCP posts - 0 post(s) |
Mr Kidd
Center for Advanced Studies Gallente Federation
70
|
Posted - 2011.10.16 18:08:00 -
[1] - Quote
I have a robotics setup that has me baffled. For the past month since setup the advanced processors producing the robotics seem to be idle half the time I look in on them which is frequent considering this is a pure production planet.
My setup consists of 4 robotics adv. procs., 8 mech. parts adv. procs. & 8 con.elects. adv. procs. So, if I have this right then as long as the mech. parts and con. elect. processors are in full production why then am I having ideal robotics processors?
Everything is routed. All mechanical parts and consumer electronics get routed to storage facilities. From there, they're routed to robotics processors. Robotics gets routed back to storage. Link utilization is under 100%.
The only thing I'm concerned that may be the cause, but I don't know why, is linking. From the starport a link goes to a processor, from that processor to another and from there to another. I'm wondering if there is some unknown congestion along the way?? We want breast augmentations and sluttier clothing in the NeX! |
Velicitia
Open Designs
0
|
Posted - 2011.10.16 18:20:00 -
[2] - Quote
probably. ideally you should be routing to some form of storage facility after each processor to take care of any timing issues that there might be... |
Dean Austrene
Deep Space Shai Hulud
1
|
Posted - 2011.10.17 10:25:00 -
[3] - Quote
Mr Kidd wrote:I have a robotics setup that has me baffled. For the past month since setup the advanced processors producing the robotics seem to be idle half the time I look in on them which is frequent considering this is a pure production planet.
My setup consists of 4 robotics adv. procs., 8 mech. parts adv. procs. & 8 con.elects. adv. procs. So, if I have this right then as long as the mech. parts and con. elect. processors are in full production why then am I having ideal robotics processors?
Everything is routed. All mechanical parts and consumer electronics get routed to storage facilities. From there, they're routed to robotics processors. Robotics gets routed back to storage. Link utilization is under 100%.
The only thing I'm concerned that may be the cause, but I don't know why, is linking. From the starport a link goes to a processor, from that processor to another and from there to another. I'm wondering if there is some unknown congestion along the way??
After each process route your items to a storage facility/spaceport and re-route to next factory from there.
You may be losing produce if routed directly from factory to factory. If this is the case then the end factory will cease production while it waits for more produce incoming.
|
pussnheels
Vintage heavy industries
108
|
Posted - 2011.10.17 10:38:00 -
[4] - Quote
all of the above plus how many factoriers are producing electronic parts and mechanical parts
If i am not mistaken they each produce 20 units but your robotic factory needs 40 of each
So could be it is waiting for more parts I do not agree with what you are saying , but i will defend to the death your right to say it...... Voltaire |
Mr Kidd
Center for Advanced Studies Gallente Federation
70
|
Posted - 2011.10.17 14:09:00 -
[5] - Quote
One factory isn't dependent on the other factory's product, they're just transiting the items back to spaceport or storage and from there it gets routed to the next step in processing. Would this still be susceptible to timing issues in routing?
The only reason processors are linked to each other is to save on link costs since shorter links cost less in cpu/pg.
We want breast augmentations and sluttier clothing in the NeX! |
Velicitia
Open Designs
5
|
Posted - 2011.10.17 14:32:00 -
[6] - Quote
Mr Kidd wrote:One factory isn't dependent on the other factory's product, they're just transiting the items back to spaceport or storage and from there it gets routed to the next step in processing. Would this still be susceptible to timing issues in routing?
The only reason processors are linked to each other is to save on link costs since shorter links cost less in cpu/pg.
sort of, but not in the way you're thinking it would.
Say you've got two BIF feeding into an AIF making the Mechanical Parts
BIF 1 makes the one set of metals BIF 2 makes the other set
you're feeding the BIF from a spaceport, so in THEORY you're pulling at exactly the same time ... but in reality BIF 1 might pull at 00:01 in the timer, and 2 might pull at 00:02 (i.e. off by a second).
your AIF has enough space to hold onto the stuff, and handle the fact that things are "off" by a second or two... but eventually one of the BIF will tick "too soon" and you've just lost a full cycle of whichever material the AIF needed (as the AIF didn't have any available storage). Now that you're off by 30 minutes (when the BIF will tick again), you'll lose at least one tick from the other BIF.
The better routing scheme is to have it essentially storage (launchpad) -> factory -> storage 2 -> factory -> storage 3(launchpad)
"storage 2" can be the same launchpad as "storage 3", since you're only holding overflow for a few seconds before the next factory in the chain finishes its cycle and pulls the 20-40 units that are sitting idle in that launchpad. |
Sola Mercury
Republic Military School Minmatar Republic
0
|
Posted - 2011.10.17 20:54:00 -
[7] - Quote
No idea, sorry. But I would take a look to a single adv. processor that isn't working. Does it have its raw materials stocked? well most likely not. If not, check the routes to that adv. proc. Do the source of raw materials have stock?
May be, you route multiple sources to one consumer. I would avoid this. |
Leonard Dukes
Arbitrage Holdings Corporation
2
|
Posted - 2011.10.17 21:36:00 -
[8] - Quote
If you're certain that the Mechanical Parts and Consumer Electronics are producing at full tilt, then I concur with Sola: verify that you don't have duplicate routes going to the same Robotics processor; any excess will be wasted and result in idle processors. If all of your routes check out, then double check your math for inputs. |
Tau Cabalander
Retirement Retreat
71
|
Posted - 2011.10.17 22:37:00 -
[9] - Quote
If you have: * Processors feeding other processors * Production of inputs = consumption of inputs * Multiple consumers then you will always have idle times.
The problem I describe is because one processor will try to fill its queue completely first, which starves another processor.
There are two solutions to this sort of problem: A. Prime the processors and always keep their input queues full. B. Have production of inputs > than consumption.
|
|
|
|
Pages: [1] :: one page |
First page | Previous page | Next page | Last page |