Hacker News new | past | comments | ask | show | jobs | submit login




Well, not requiring OIDs feels like a benefit grin

(I actually understand completely why SNMP uses OIDs, and why they actually have advantages, but man is it tough to get into SNMP)


Complex joining of multiple data sets with filters and conditions utilizing a query language that has been utilized to great effect for longer than SNMP has existed.


The output of snmptable is already hierarchal and there is already well-defined MIB for thousands of data sources.

I guess using SQL instead of regex is the benefit you mean? Certainly more readable.

Not that it really matters much to conversation at hand but standard SQL is not really that much older than SNMP. Both were late 80s if memory serves.


Truthfully, I'm not entirely sure the capabilities of snmptable, but from your statement I assume it allows joining of different sets of data in ways to show specific correlated data (I'm not seeing that when searching, but I'll assume my search skills are to blame here). In that case, the real benefit of this is SQL the language, which while it would be overkill if it was designed for just this application, is well understood and utilized and a has a large existing userbase.

I'm not sure how to use snmptable, as I've never had the need, and while I suspect in the simple cases it may be easier to learn and use than SQL, I'm highly skeptical the medium complexity cases that SQL handles are even possible in it (unless it has a fairly well fleshed out query language of it's own, as opposed to specifying parameters on a command line).


About 15 years apart to be exact: 1974 for SQL http://www.almaden.ibm.com/cs/people/chamberlin/sequel-1974.... and 1988 for SNMP http://tools.ietf.org/html/rfc1065 .


shawn-butler is technically correct -- he refers to standard SQL (first standardized in 1986) and SNMP being about the same time, and both in the late 1980s, and this is correct.

Of course, SQL was around quite a while before being standardized.


I looked it up before posting, and saw the same thing, that SQL was standardized the year before the SNMP RFCs were put out, but that it was in use since the 70s. That made the statement correct by either a little or a lot, depending on how you wanted to look at it, so I felt safe in making it.




Consider applying for YC's W25 batch! Applications are open till Nov 12.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: