.. etwork). (Stallings, 213) The largest problem with SNMP, ironically enough, is the same thing that made it great; its simple design. Because it is so simple, the information it deals with is neither detailed, nor well organized enough to deal with the growing networks of the 1990s. This is mainly due to the quick creation of SNMP, because it was never designed to be the network management protocol of the 1990s.
Like the previous flaw, this one too has been corrected with the new version, SNMPv2. This new version allows for more in-detail specification of variables, including the use of the table data structure for easier data retrieval. Also added are two new PDUs that are used to manipulate the tabled objects. In fact, so many new features have been added that the formal specifications for SNMP have expanded from 36 pages (with v1) to 416 pages with SNMPv2. (Stallings, 153) Some people might say that SNMPv2 has lost the simplicity, but the truth is that the changes were necessary, and could not have been avoided.
A management station relies on the agent at a device to retrieve or update the information at the device. The information is viewed as a logical database, called a Management Information Base, or MIB. MIB modules describe MIB variables for a large variety of device types, computer hardware, and software components. The original MIB for Managing a TCP/IP internet (now called MIB-I) was defined in RFC 1066 in August of 1988. It was updated in RFC 1156 in May of 1990.
The MIB-II version published in RFC 1213 in May of 1991, contained some improvements, and has proved that it can do a good job of meeting basic TCP/IP management needs. MIB-II added many useful variables missing from MIB-I (Feit, 85). MIB files are common variables used not only by SNMP, but CMIP as well. In the late 1980s a project began, funded by governments, and large corporations. Common Management Information Protocol (CMIP) was born. Many thought that because of its nearly infinite development budget, that it would quickly become in widespread use, and overthrow SNMP from its throne. Unfortunately, problems with its implementation have delayed its use, and it is now only available in limited form from developers themselves.
(SNMP, Part 2 of 2, III.40.) CMIP was designed to be better than SNMP in every way by repairing all flaws, and expanding on what was good about it, making it a bigger and more detailed network manager. Its design is similar to SNMP, where PDUs are used as variables to monitor the network. CMIP however contains 11 types of PDUs (compared to SNMPs 5). In CMIP, the variables are seen as very complex and sophisticated data structures with three attributes. These include: 1) Variable attributes: which represent the variables characteristics (its data type, whether it is writable) 2) variable behaviors: what actions of that variable can be triggered.
3) Notifications: the variable generates an event report whenever a specified event occurs (eg. A terminal shutdown would cause a variable notification event) (Comer, 82) As a comparison, SNMP only employs variable properties from one and three above. The biggest feature of the CMIP protocol is that its variables not only relay information to and from the terminal (as in SNMP) , but they can also be used to perform tasks that would be impossible under SNMP. For instance, if a terminal on a network cannot reach the fileserver a pre-determined amount of times, then CMIP can notify appropriate personnel of the event. With SNMP however, a user would have to specifically tell it to keep track of unsuccessful attempts to reach the server, and then what to do when that variable reaches a limit.
CMIP therefore results in a more efficient management system, and less work is required from the user to keep updated on the status of the network. CMIP also contains the security measures left out by SNMP. Because of the large development budget, when it becomes available, CMIP will be widely used by the government, and the corporations that funded it. After reading the above paragraph, you might wonder why, if CMIP is this wonderful, is it not being used already? (after all, it had been in development for nearly 10 years) The answer is that possibly CMIPs only major disadvantage, is enough in my opinion to render it useless. CMIP requires about ten times the system resources that are needed for SNMP.
In other words, very few systems in the world would able to handle a full implementation on CMIP without undergoing massive network modifications. This disadvantage has no inexpensive fix to it. For that reason, many believe CMIP is doomed to fail. The other flaw in CMIP is that it is very difficult to program. Its complex nature requires so many different variables that only a few skilled programmers are able to use it to its full potential. Considering the above information, one can see that both management systems have their advantages and disadvantages. However the deciding factor between the two, lies with their implementation, for now, it is almost impossible to find a system with the necessary resources to support the CMIP model, even though it is superior to SNMP (v1 and v2) in both design and operation. Many people believe that the growing power of modern systems will soon fit well with CMIP model, and might result in its widespread use, but I believe by the time that day comes, SNMP could very well have adapted itself to become what CMIP currently offers, and more.
As weve seen with other products, once a technology achieves critical mass, and a substantial installed base, its quite difficult to convince users to rip it out and start fresh with an new and unproven technology (Borsook, 48). It is then recommend that SNMP be used in a situation where minimial security is needed, and SNMPv2 be used where security is a high priority. Bibliography “SNMP tools evolving to meet critical LAN needs.” Infoworld June 1, 1992: 48-49. Comer, Douglas E. Internetworking with TCP/IP New York: Prentice-Hall, Inc., 1991. Dryden, Partick. “Another view for SNMP.” Computerworld December 11, 1995: 12.
Feit, Dr. Sidnie. SNMP. New York: McGraw-Hill Inc., 1995. Fisher, Sharon.
“Dueling Protocols.” Byte March 1991: 183-190. Horwitt, Elisabeth. “SNMP holds steady as network standard.” Computerworld June 1, 1992: 53-54. Leon, Mark. “Advent creates Java tools for SNMP apps.” Infoworld March 25, 1996: 8. Marshall, Rose.
The Simple Book. New Jersey: Prentice Hall, 1994. Miller, Mark A., P.E. Managing Internetworks with SNMP New York: M&T Books, 1993. Moore, Steve. “Committee takes another look at SNMP.” Computerworld January 16, 1995: 58.
Moore, Steve. “Users weigh benefits of DMI, SNMP.” Computerworld July, 31 1995: 60. The SNMP Workshop & Panther Digital Corporation. SNMP FAQ Part 1 of 2. Danbury, CT: http://www.www.cis.ohio-state.edu/hypertext/faq/us enet/snmp- faq/part1/faq.html, . The SNMP Workshop & Panther Digital Corporation. SNMP FAQ Part 2 of 2.
Danbury, CT: http://www.www.cis.ohio-state.edu/hypertext/faq/us enet/snmp- faq/part2/faq.html, . Stallings, William. SNMP, SNMPv2, and CMIP. Don Mills, Addison-Wesley, 1993. Vallillee, Tyler, web page author.
Http://www.undergrad.math. uwaterloo.ca/~tkvallil/snmp.html VanderSluis, Kurt. “SNMP: Not so simple.” MacUser October 1992: 237-240.