1. The method of making a printed wiring board without the formation of slivers, comprising the steps of:
a. providing a substrate having a plurality of thin circuit lines thereon;
b. applying a solder mask over the circuit lines onto which a commoning layer is not to be applied;
c. applying a seed catalyst layer over the solder mask and over all surfaces of the substrate that are not covered by the solder mask;
d. applying a thin metal commoning layer of electroless copper on the substrate and on the circuit lines to a thickness sufficient to achieve electrical continuity;
e. applying a photoresist over the thin commoning layer;
f. removing the photoresist to expose only the portion of the commoning layer covering the circuit lines as well as the side walls of the circuit lines; and
g. electrodepositing a conforming layer of a metal more noble than copper over the exposed commoning layer, covering the top and extending down the sides of each circuit line, at least to the point at which the conforming layer of electroless copper is intersected by the more noble metal.
2. The method according to claim 1 wherein the more noble metal electrodeposited over the exposed commoning layer is selected from the group consisting of zinc, cadmium, chromium, nickel, cobalt, gold, silver, palladium, platinum, ruthenium, and alloys with each other and with tin and lead.
3. The method according to claim 2 wherein the more noble metal comprises an electrodeposit of gold over an electrodeposit of nickel.
4. The method according to claim 3 wherein the cross-sectional shape of the circuit lines and commoning layer thereon are in the form of a trapezoid or a rectangle.
5. The method according to claim 4 wherein the nickelgold forms a conforming layer on the top and the sides of the circuit trace and commoning layer onto which they are electroplated.
6. The method according to claim 1 wherein the photoresist is removed after deposition of the conforming layer, followed by removal of all of the exposed portions of the commoning layer.
7. The method according to claim 6 wherein one or more surface finish metallization layers may be added by patterning additional layers of photoresist before removal of the exposed commoning layer.
8. The method according to claim 1 wherein the thin circuit lines are deposited on the surface of the printed wiring board by the steps of:
a) adhering a copper foil to the surface of the printed wiring board, and
b) electrodepositing a layer of copper over the foil.
9. The method according to claim 8 further including forming a flash copper layer on the copper layer prior to electrodepositing said layer of copper.
10. The method according to claim 8 wherein unwanted portions of the copper foil are removed by etching prior to electrodepositing said layer of copper over the remaining foil.
11. The method according to claim 1 wherein the commoning layer is deposited to a thickness of between about 15\u03bc inches and about 200\u03bc inches on the substrate and the exposed circuit lines.
12. The method of manufacturing a printed wiring board having a circuitized metal electrodeposit on both sides thereof, comprising the steps of
a. forming at least one continuous metallization layer of copper on each side of a dielectric substrate;
b. creating circuit traces on each metallization layer by removing unwanted metal;
c. masking over any area of any circuit trace that does not require a selective layer of a more noble metal;
d. applying a seed catalyst layer over all masked and unmasked surfaces;
e. flash plating a commoning layer of copper to a thickness between about 15\u03bc inches and about 200\u03bc inches over the seed catalyst layer;
f. applying a photoresist film over the commoning layer;
g. removing the photoresist film to expose the commoning layer over circuit traces to be plated;
h. electrolytically plating a conforming layer of the more noble metal than that of the commoning layer over the exposed commoning layer, covering the top and extending down the sides of each circuit line, at least to the point at which the conforming layer is intersected by the more noble metal;
i. stripping the photoresist; and
j. removing the commoning layer between the circuit traces.
13. The method according to claim 12 wherein the seed catalyst is a tinpalladium catalyst.
14. The method according to claim 12 further including the step of roughening the surface of the mask before applying the seed catalyst.
15. The method according to claim 12 wherein the commoning layer is exposed over the circuit traces to be plated by exposing and developing the photoresist film with a laser beam or a collimated light source to create openings through the film to the commoning layer.
16. The method according to claim 12 wherein the more noble metal is selected from the group consisting of zinc, cadmium, chromium, nickel, cobalt, gold, silver, palladium, platinum, ruthenium, and alloys with each other and with tin and lead.
17. The method according to claim 12 wherein the circuit traces are formed by the additional step of electrodepositing a layer of copper on the created circuit traces prior to masking of the circuit traces.
18. The method according to claim 17 further including forming a flash copper layer over the circuit traces prior to electrodepositing the copper layer over the circuit traces.
The claims below are in addition to those above.
All refrences to claim(s) which appear below refer to the numbering after this setence.
What is claimed is:
1. A method for reestablishing connection information on a switch connected to plural servers in a computer network, comprising:
identifying, at the switch, information to be transferred through the switch between a client and a first server;
determining an absence of connection information on the switch used to transfer the information between the client and the first server through the switch; and
responding to the determining by reconstructing the connection information on the switch using connection information extracted from an information packet passing through the switch from at least one of the plural servers during a communication associated with the information.
2. The method of claim 1, wherein the information to be transferred through the switch is information included in the information packet from the at least one of the plural servers.
3. The method of claim 1, wherein the information to be transferred through the switch is information included in a first information packet transferred to the switch from the client, and the connection information is reconstructed on the switch using a second information packet sent by the at least one server to the switch in response to the first information packet.
4. The method of claim 1, comprising:
determining a group membership of the computer network, wherein the group membership includes the plural servers that are active in the computer network.
5. The method of claim 4, wherein the group membership of the computer network is determined at predetermined time intervals.
6. The method of claim 4, comprising:
allocating, by the plural servers, a computer network connection address space among the plural servers.
7. The method of claim 6, wherein the computer network connection address space is allocated among the plural servers based upon a connection load of each of the plural servers.
8. The method of claim 6, wherein at least one of the plural servers notifies the switch of a dispersal algorithm be used by the switch for allocating computer network connection address space of the plural servers.
9. The method of claim 8, wherein upon determining an absence of connection information on the switch, the information is transferred to a victim server using the dispersal algorithm.
10. The method of claim 9, wherein the dispersal algorithm is a predetermined matching function which is known to the plural servers and which is stored at the switch.
11. The method of claim 10, comprising:
storing the information in a computer memory of the victim server; and
forwarding the information packet from the victim server to the at least one server.
12. The method of claim 11, wherein the information packet is forwarded through the switch from the victim server to the at least one server, and wherein the information packet contains the address of the at least one server.
13. The method of claim 1, wherein the connection information is a connection tuple of the information packet.
14. The method of claim 13, wherein the responding comprises:
extracting, by the switch, the connection tuple from the information packet on the switch.
15. A system for reestablishing connection information in a switch of a computer network, comprising:
plural servers;
a switch, connected to each of the plurality of servers, for identifying information transferred through the switch between a client and a first server, for determining an absence of connection information on the switch used to transfer the information from the client to the first server through the switch, and for responding to the determination of the absence of connection information by reconstructing the connection information on the switch using connection information extracted from an information packet passing through the switch from at least one of the plurality of servers during a communication associated with the information.
16. The method of claim 15, wherein the information to be transferred through the switch is information included in the information packet from the at least one of the plural servers.
17. The method of claim 15, wherein the information to be transferred through the switch is information included in a first information packet transferred to the switch from the client, and the connection information is reconstructed on the switch using a second information packet sent by the at least one server to the switch in response to the first information packet.
18. The system of claim 15, wherein the switch and the plurality of servers determine a group membership of the computer network, and wherein the group membership includes the switch and the plurality of servers that are active in the computer network.
19. The system of claim 18, wherein the group membership of the computer network is determined at predetermined time intervals.
20. The system of claim 18, wherein the plurality of servers allocate a computer network connection address space among the plurality of servers.
21. The system of claim 18, wherein at least one of the plural servers notifies the switch of a dispersal algorithm to be used by the switch for allocating computer network connection address space of the plurality of servers.
22. The system of claim 21, wherein the dispersal algorithm is a predetermined matching function which is known to be plurality of servers and which is stored at the switch.
23. The system of claim 15, wherein the connection information is a connection tuple of the information packet.