Regulation of traffic in corporate IM networks.
By the nature of work, I had the opportunity to implement IM (Instant Messenger) within the company. Yes, not one, but a network of three. About the implementation of IM, mechanisms for collecting and counting traffic, choosing the right client is a separate article or even three. I want to start a conversation about the regulation of traffic in such networks. The company only threw out about 10 thousand dollars to limit HTTP traffic in the company and here you have the bosses about Instant Messenger. Of course, convenient and progressive, but this is the most time killer. There was a question that remains open now - how to prevent abuse of the service.
It is no secret that the administrative method of influencing users is the most effective. The very fact that traffic is under control and someone can read this is already a good reason to restrict chatter. Especially if the bosses in the office are not a rag, and it’s not a question for a couple of good exponential scoldings or even layoffs.
Collecting all the traffic in the database and the web interface to view the traffic and statistics was thrown to the authorities for a couple of days. But the boss is the boss, and the king in the head is different. The idea was born to send a “letter-cap” to lovers to chat every morning with a text like this (free translation from English):
It seems to be in a measure of intimidation and formality (forgive my translation, but for some reason I get better English messages). It remains to determine who exactly.
Initially, the idea was as follows - all with the number of received + sent messages are above average. But such turned out to be quite a lot. The people began to complain about the extra traffic - it is clear to everyone that this whining is mostly contrived. But it made me think.
All users with the number of messages less than the daily maximum divided by 10 were ejected from the sample. if the most flooding wrote / received 1500 messages, then all with modest figures of 150 and below will not be taken into account. But still it was wrong. One such flooder could throw out all of the sample. The last option was Top 10. That is. take the ten most presamy and warn them.
And then the authorities called and said, "And let's stop sending out reminders for two months and see what happens" (read, "and fire a couple of people"). Waiting for orders to cover the service;)
It is no secret that the administrative method of influencing users is the most effective. The very fact that traffic is under control and someone can read this is already a good reason to restrict chatter. Especially if the bosses in the office are not a rag, and it’s not a question for a couple of good exponential scoldings or even layoffs.
Collecting all the traffic in the database and the web interface to view the traffic and statistics was thrown to the authorities for a couple of days. But the boss is the boss, and the king in the head is different. The idea was born to send a “letter-cap” to lovers to chat every morning with a text like this (free translation from English):
Dear user. This is a reminder that all traffic is logged and can be viewed. Please use this service for business purposes only.
')
According to yesterday's statistics, you sent X and received X messages (X total). This means that with an average message length of 35 characters and a reading or writing speed of 120 characters per minute, you spent X minutes using this service. We hope that all messages were for work.
It seems to be in a measure of intimidation and formality (forgive my translation, but for some reason I get better English messages). It remains to determine who exactly.
Initially, the idea was as follows - all with the number of received + sent messages are above average. But such turned out to be quite a lot. The people began to complain about the extra traffic - it is clear to everyone that this whining is mostly contrived. But it made me think.
All users with the number of messages less than the daily maximum divided by 10 were ejected from the sample. if the most flooding wrote / received 1500 messages, then all with modest figures of 150 and below will not be taken into account. But still it was wrong. One such flooder could throw out all of the sample. The last option was Top 10. That is. take the ten most presamy and warn them.
And then the authorities called and said, "And let's stop sending out reminders for two months and see what happens" (read, "and fire a couple of people"). Waiting for orders to cover the service;)
Source: https://habr.com/ru/post/15834/
All Articles