tip: web requests should not be measured in Hz [hertz] as that is only used for periodic frequencies, which random events (like requests hitting a web server) are not!
-
@sophie good idea, are they also poisson distributed?
@Quantensalat @sophie Don't open that can of worms. In university I had an entire compulsory 1-semester course on queueing theory.
Rabbit hole exit node: https://en.wikipedia.org/wiki/Erlang_distribution
-
tip: web requests should not be measured in Hz [hertz] as that is only used for periodic frequencies, which random events (like requests hitting a web server) are not!
measure them in Bq [becquerel] instead
-
@4censord @sophie I remember talking to someone in the late 90s, early 00s that told me a colleague had tied in a sound generator to their company’s smtp servers, and it would play forest sounds in the background all day in the sysadmin office. I seem to recall that the amount of rain was tied to the load, and different bird calls represented different types and sizes of mail.
It was done in such a way as to be a pleasant background sound, but at the same time, when something went wrong, the sysops would hear it long before monitoring flagged it.
I suspect this system eventually died a death due to moving to a cloud provider, but it does show that monitoring can be something other than visual…
-
tip: web requests should not be measured in Hz [hertz] as that is only used for periodic frequencies, which random events (like requests hitting a web server) are not!
measure them in Bq [becquerel] instead
@sophie@mastodon.catgirl.cloud what if the requests are periodic tho?
-
@sophie and use clicks to indicate each one!
geiger counter here we gooo!! -
tip: web requests should not be measured in Hz [hertz] as that is only used for periodic frequencies, which random events (like requests hitting a web server) are not!
measure them in Bq [becquerel] instead
@sophie
SI avoids identifying countable entities and events like web requests except in special cases like becquerel (which is specifically for radioactive decay). This is but one example among many where the needs of IT are outside the scope that SI serves.
https://www.researchgate.net/publication/339954122_Quantities_and_Units_for_Software_Product_Measurements -
tip: web requests should not be measured in Hz [hertz] as that is only used for periodic frequencies, which random events (like requests hitting a web server) are not!
measure them in Bq [becquerel] instead
@sophie the normal unit for these kinds of things is the Erlang, but you would have to normalize for the number of requests/second per active user.
Perhaps create a new unit? I vot for BernersLee, abbreviation BL.
-
@sophie the normal unit for these kinds of things is the Erlang, but you would have to normalize for the number of requests/second per active user.
Perhaps create a new unit? I vot for BernersLee, abbreviation BL.
@fazalmajid @sophie isn't erlang only applicable to circuit-switched networks like analog telephone, not for packet-switched networks like TCP/IP?
-
@fazalmajid @sophie isn't erlang only applicable to circuit-switched networks like analog telephone, not for packet-switched networks like TCP/IP?
-
@sophie
SI avoids identifying countable entities and events like web requests except in special cases like becquerel (which is specifically for radioactive decay). This is but one example among many where the needs of IT are outside the scope that SI serves.
https://www.researchgate.net/publication/339954122_Quantities_and_Units_for_Software_Product_MeasurementsBq is fine. Network engineers have come to call the unavoidable botnet probing a system will see within minutes after being connected as the background radiation of the Internet.
-
tip: web requests should not be measured in Hz [hertz] as that is only used for periodic frequencies, which random events (like requests hitting a web server) are not!
measure them in Bq [becquerel] instead
@sophie practically, however, both end up being interpreted as the same
-
tip: web requests should not be measured in Hz [hertz] as that is only used for periodic frequencies, which random events (like requests hitting a web server) are not!
measure them in Bq [becquerel] instead
-
@Quantensalat @sophie Don't open that can of worms. In university I had an entire compulsory 1-semester course on queueing theory.
Rabbit hole exit node: https://en.wikipedia.org/wiki/Erlang_distribution
@hennichodernich @sophie oh wow, but I should have guessed that its a crucial topic
-
@4censord @sophie I remember talking to someone in the late 90s, early 00s that told me a colleague had tied in a sound generator to their company’s smtp servers, and it would play forest sounds in the background all day in the sysadmin office. I seem to recall that the amount of rain was tied to the load, and different bird calls represented different types and sizes of mail.
It was done in such a way as to be a pleasant background sound, but at the same time, when something went wrong, the sysops would hear it long before monitoring flagged it.
I suspect this system eventually died a death due to moving to a cloud provider, but it does show that monitoring can be something other than visual…
@moof But wouldn't it be more prone to just rainy day exploits, upon cloud integration?
-
@4censord @sophie I remember talking to someone in the late 90s, early 00s that told me a colleague had tied in a sound generator to their company’s smtp servers, and it would play forest sounds in the background all day in the sysadmin office. I seem to recall that the amount of rain was tied to the load, and different bird calls represented different types and sizes of mail.
It was done in such a way as to be a pleasant background sound, but at the same time, when something went wrong, the sysops would hear it long before monitoring flagged it.
I suspect this system eventually died a death due to moving to a cloud provider, but it does show that monitoring can be something other than visual…
-
@4censord ah yes, the prometheus→grafana→geiger counter monitoring stack, who doesn't love it
-
@4censord @sophie I remember talking to someone in the late 90s, early 00s that told me a colleague had tied in a sound generator to their company’s smtp servers, and it would play forest sounds in the background all day in the sysadmin office. I seem to recall that the amount of rain was tied to the load, and different bird calls represented different types and sizes of mail.
It was done in such a way as to be a pleasant background sound, but at the same time, when something went wrong, the sysops would hear it long before monitoring flagged it.
I suspect this system eventually died a death due to moving to a cloud provider, but it does show that monitoring can be something other than visual…
-
@4censord @sophie I remember talking to someone in the late 90s, early 00s that told me a colleague had tied in a sound generator to their company’s smtp servers, and it would play forest sounds in the background all day in the sysadmin office. I seem to recall that the amount of rain was tied to the load, and different bird calls represented different types and sizes of mail.
It was done in such a way as to be a pleasant background sound, but at the same time, when something went wrong, the sysops would hear it long before monitoring flagged it.
I suspect this system eventually died a death due to moving to a cloud provider, but it does show that monitoring can be something other than visual…
@moof @4censord @sophie
One of my jobs during my “I’m contemplating my third switch of academic majors, perhaps I should take some time off from school to think about this” wanderjahren was babysitting the larval Internet at night.(push digression)
I would call the phone company and complain about high-bandwidth (a whole 56K!) phone lines. Occasionally I would have to disturb a night-watchman and talk them through looking at the light panel on the Interface Message Processor (IMP — larval switches or gateways in today’s thinking). Very rarely I’d talk them through toggling the 16-word boot loader that would boot the IMP from a neighbor through the modem.
(pop digression)Logs were printed by an inkjet printer — silent, save for line-feeds.
I could hear certain patterns of line feeds and (coupled with the time of day and time of year) know which IMP needed help without looking at the log before the monitoring host alarm timeout.
(push digression)
Time of day and time of year? What?In the southern US during those days, many phone lines were carried by microwave — I’m sure the towers for these have all been replaced by cell towers (more likely fiber buried along a rail line).
During the spring, as the sun rose, the damp would rise from the rivers and lakes. The mist would interfere with the microwaves, and I could watch in the log as the sun rose and phone lines failed in a line from east to west.
(pop digression)This pattern also had a characteristic pattern of line-feeds.
(push to possibly apocryphal digression)
That’s not the only meteorological phenomenon visible in network traffic logs. I heard it said that David Mills, the creator of the Network Time Protocol (NTP), could tell when a heat-wave hit the American Midwest, because the sun would heat the copper wires carrying phone signals, they’d expand, and the altered distance across the United States would show up in NTP packet timing.
(pop from digression) -
@4censord @sophie I remember talking to someone in the late 90s, early 00s that told me a colleague had tied in a sound generator to their company’s smtp servers, and it would play forest sounds in the background all day in the sysadmin office. I seem to recall that the amount of rain was tied to the load, and different bird calls represented different types and sizes of mail.
It was done in such a way as to be a pleasant background sound, but at the same time, when something went wrong, the sysops would hear it long before monitoring flagged it.
I suspect this system eventually died a death due to moving to a cloud provider, but it does show that monitoring can be something other than visual…
-
R relay@relay.infosec.exchange shared this topicR relay@relay.mycrowd.ca shared this topic
-
@4censord @sophie I remember talking to someone in the late 90s, early 00s that told me a colleague had tied in a sound generator to their company’s smtp servers, and it would play forest sounds in the background all day in the sysadmin office. I seem to recall that the amount of rain was tied to the load, and different bird calls represented different types and sizes of mail.
It was done in such a way as to be a pleasant background sound, but at the same time, when something went wrong, the sysops would hear it long before monitoring flagged it.
I suspect this system eventually died a death due to moving to a cloud provider, but it does show that monitoring can be something other than visual…
since it also goes into more detail.
