observium managment should be more professional
http://afmug.com/pipermail/af/2015-March/018914.html
As a provider of internet connectivity (ergo a telco and a ISP) we use licensed microwave and unlicensed microwave technologies to feed our revenue generating customers. We also use fiber, and copper based technologies.
microwave != wifi
We are not alone in using a mix of technologies to deliver bits to customers. T-Mobile, ATT, Sprint, AllTel, Cellular One, Verizon, etc all use microwave (wireless) technologies to deliver bits.
I, like others, have offered to write code, have asked for pointers on how to get started writing code, all appears to have fallen upon deaf ears.
I've offered to PAY others to write code to support our kit, with features we need. No takers.
As a PAID user, I'm quite shocked and disappointed in Adam's childish behaviour.
Users of Observium should be concerned that their continued investment could be a loosing path as more people move AWAY from the product and go support other packages. On two lists that I saw the above URL, not less than 3 dozen people said they where removing observium and not renewing their annual subscriptions. I suspect that will become a trend.
Cheers
John,
When you consider how complex it is to "write software" its not something you can take up gradually. A Google fellow ALL OF 32, estimates in his lifetime he has spent 50,000 hours writing software. I'm guessing that is 20 years, perhaps 60 hours a week (since he was a kid), as a "year" only has 8,760 hours.
And even Google the other day had massive gmail outage that CNBC reported during their telecast despite 1000's of pro's like him on staff tasked with their "software".
It's not some nothing you take up in your spare time and have much luck at it, imho.
If the cursory graphs and checks (which are pretty nice) in Observium are not enough, I recommend brining in IBM netcool or HP Openview teams like many major ISP's (Level3, ATT) have had to do. Of course, these consultants rate's start at $250/hour in my experience and not $100 a year like Observium :)
In our similar operation to your own, ran on behalf of private low latency trading firms between NY and CHICAGO, with extremely secretive, precise requirements - we monitor the "overall" network with Observium to get "a sense" of utilization at 20,000 feet - then drill down with very very precise tools written for each purpose in python, java, bash shells, etc - each with its own backend data storage and reporting scheme from a simple as rrd files, log files, pcap files to as complex as Microsoft sql and couchbase.
In short, I think you are trying to do too much with one tool. You need a "polling style" app for the overview time series, and you need "real-time sensors" that trip and alert in a second or less. We use many variants of both. I call it "monitoring in depth". So even if 3 out of 4 of our tools fail - the 4th one will alert us and the client, often with seconds to "go flat". Imagine you were writing call options in microseconds and you didn't have the current stock price or news about the stock :)
Thanks,
Joe
-----Original Message----- From: observium [mailto:observium-bounces@observium.org] On Behalf Of John Brown Sent: Saturday, April 04, 2015 05:56 To: observium Subject: [Observium] observium managment should be more professional
http://afmug.com/pipermail/af/2015-March/018914.html
As a provider of internet connectivity (ergo a telco and a ISP) we use licensed microwave and unlicensed microwave technologies to feed our revenue generating customers. We also use fiber, and copper based technologies.
microwave != wifi
We are not alone in using a mix of technologies to deliver bits to customers. T-Mobile, ATT, Sprint, AllTel, Cellular One, Verizon, etc all use microwave (wireless) technologies to deliver bits.
I, like others, have offered to write code, have asked for pointers on how to get started writing code, all appears to have fallen upon deaf ears.
I've offered to PAY others to write code to support our kit, with features we need. No takers.
As a PAID user, I'm quite shocked and disappointed in Adam's childish behaviour.
Users of Observium should be concerned that their continued investment could be a loosing path as more people move AWAY from the product and go support other packages. On two lists that I saw the above URL, not less than 3 dozen people said they where removing observium and not renewing their annual subscriptions. I suspect that will become a trend.
Cheers _______________________________________________ observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
Joe, as a person that has written millions of lines of code, starting in the mid-seventies with BASIC, Fortran (both 4 and 77), Pascal, ASM (6502, Z80, 68K, 8080, x86, SPARC, MIPS, etc), C and some C++, along with scripting languages like PHP, Perl, and Python, I can say with reasonable certainty that what is being asked is not more complex or harder than what has already been produced.
Certainly being a jerk, using cus words and generally being highly unprofessional has zip to do with actually writing code. It does have everything to with being an ass.
What the leader of Observium has clearly said is that an entire market segment and technology type is not worth his time and he has done so in a most unprofessional manner. There have been MANY people, myself included, that have been willing to compensate by way of money, spend our time helping develop code, etc, all to be blown off and cus'ed at.
Maybe LibreNMS will be more successful since its a community project, of people, by people and for the community.
Hundreds of people have asked for the ability to add the ability to graph, alert, etc on a OID. Not an unreasonable request. A custom sensor or similar could be created. The customer base has made requests, the keeper of the keys has told them to pound sand.
On Sat, Apr 4, 2015 at 5:27 PM, Joseph L. Brunner joe@affirmedsystems.com wrote:
John,
When you consider how complex it is to "write software" its not something you can take up gradually. A Google fellow ALL OF 32, estimates in his lifetime he has spent 50,000 hours writing software. I'm guessing that is 20 years, perhaps 60 hours a week (since he was a kid), as a "year" only has 8,760 hours.
And even Google the other day had massive gmail outage that CNBC reported during their telecast despite 1000's of pro's like him on staff tasked with their "software".
It's not some nothing you take up in your spare time and have much luck at it, imho.
If the cursory graphs and checks (which are pretty nice) in Observium are not enough, I recommend brining in IBM netcool or HP Openview teams like many major ISP's (Level3, ATT) have had to do. Of course, these consultants rate's start at $250/hour in my experience and not $100 a year like Observium :)
In our similar operation to your own, ran on behalf of private low latency trading firms between NY and CHICAGO, with extremely secretive, precise requirements - we monitor the "overall" network with Observium to get "a sense" of utilization at 20,000 feet - then drill down with very very precise tools written for each purpose in python, java, bash shells, etc - each with its own backend data storage and reporting scheme from a simple as rrd files, log files, pcap files to as complex as Microsoft sql and couchbase.
In short, I think you are trying to do too much with one tool. You need a "polling style" app for the overview time series, and you need "real-time sensors" that trip and alert in a second or less. We use many variants of both. I call it "monitoring in depth". So even if 3 out of 4 of our tools fail - the 4th one will alert us and the client, often with seconds to "go flat". Imagine you were writing call options in microseconds and you didn't have the current stock price or news about the stock :)
Thanks,
Joe
-----Original Message----- From: observium [mailto:observium-bounces@observium.org] On Behalf Of John Brown Sent: Saturday, April 04, 2015 05:56 To: observium Subject: [Observium] observium managment should be more professional
http://afmug.com/pipermail/af/2015-March/018914.html
As a provider of internet connectivity (ergo a telco and a ISP) we use licensed microwave and unlicensed microwave technologies to feed our revenue generating customers. We also use fiber, and copper based technologies.
microwave != wifi
We are not alone in using a mix of technologies to deliver bits to customers. T-Mobile, ATT, Sprint, AllTel, Cellular One, Verizon, etc all use microwave (wireless) technologies to deliver bits.
I, like others, have offered to write code, have asked for pointers on how to get started writing code, all appears to have fallen upon deaf ears.
I've offered to PAY others to write code to support our kit, with features we need. No takers.
As a PAID user, I'm quite shocked and disappointed in Adam's childish behaviour.
Users of Observium should be concerned that their continued investment could be a loosing path as more people move AWAY from the product and go support other packages. On two lists that I saw the above URL, not less than 3 dozen people said they where removing observium and not renewing their annual subscriptions. I suspect that will become a trend.
Cheers _______________________________________________ observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
All I see in that thread is even more of the same entitled behavior as the idiot who refused to take "no" for an answer for half an hour and got banned from the channel.
This ridiculous drama is based purely on a small section of our user base believing they can demand that we support the devices they use, because, well, they use them. WISPs are a tiny minority of our user base, and an even tinier minority of our paying user base.
There's simply no justification for the huge amount of effort required to build support for these devices to Observium's required standards.
Also, if you read that thread, it's hilariously clear that no one actually read it, because almost no one picks up on the fact that the original poster claims I called all WISPs retarded, when in reality I was referring to the vendors who write the MIBs (which is an undeniable fact to anyone who has ever tried to get information from wireless kit via SNMP)! :DDD
In any event, all this demonstrates to me is that this is a sector of the market which is seemingly way too much hassle to try to support.
adam.
p.s.
Best part of this email is that my mail client marked "cus" as a spelling mistake, since it isn't actually a word, and belies the puritanical scared-of-words origins of the writer.
Not everyone lives in the "must not cuss!" USA!
On 2015-04-05 02:00, John Brown wrote:
Joe, as a person that has written millions of lines of code, starting in the mid-seventies with BASIC, Fortran (both 4 and 77), Pascal, ASM (6502, Z80, 68K, 8080, x86, SPARC, MIPS, etc), C and some C++, along with scripting languages like PHP, Perl, and Python, I can say with reasonable certainty that what is being asked is not more complex or harder than what has already been produced.
Certainly being a jerk, using cus words and generally being highly unprofessional has zip to do with actually writing code. It does have everything to with being an ass.
What the leader of Observium has clearly said is that an entire market segment and technology type is not worth his time and he has done so in a most unprofessional manner. There have been MANY people, myself included, that have been willing to compensate by way of money, spend our time helping develop code, etc, all to be blown off and cus'ed at.
Maybe LibreNMS will be more successful since its a community project, of people, by people and for the community.
Hundreds of people have asked for the ability to add the ability to graph, alert, etc on a OID. Not an unreasonable request. A custom sensor or similar could be created. The customer base has made requests, the keeper of the keys has told them to pound sand.
On Sat, Apr 4, 2015 at 5:27 PM, Joseph L. Brunner joe@affirmedsystems.com wrote:
John,
When you consider how complex it is to "write software" its not something you can take up gradually. A Google fellow ALL OF 32, estimates in his lifetime he has spent 50,000 hours writing software. I'm guessing that is 20 years, perhaps 60 hours a week (since he was a kid), as a "year" only has 8,760 hours.
And even Google the other day had massive gmail outage that CNBC reported during their telecast despite 1000's of pro's like him on staff tasked with their "software".
It's not some nothing you take up in your spare time and have much luck at it, imho.
If the cursory graphs and checks (which are pretty nice) in Observium are not enough, I recommend brining in IBM netcool or HP Openview teams like many major ISP's (Level3, ATT) have had to do. Of course, these consultants rate's start at $250/hour in my experience and not $100 a year like Observium :)
In our similar operation to your own, ran on behalf of private low latency trading firms between NY and CHICAGO, with extremely secretive, precise requirements - we monitor the "overall" network with Observium to get "a sense" of utilization at 20,000 feet - then drill down with very very precise tools written for each purpose in python, java, bash shells, etc - each with its own backend data storage and reporting scheme from a simple as rrd files, log files, pcap files to as complex as Microsoft sql and couchbase.
In short, I think you are trying to do too much with one tool. You need a "polling style" app for the overview time series, and you need "real-time sensors" that trip and alert in a second or less. We use many variants of both. I call it "monitoring in depth". So even if 3 out of 4 of our tools fail - the 4th one will alert us and the client, often with seconds to "go flat". Imagine you were writing call options in microseconds and you didn't have the current stock price or news about the stock :)
Thanks,
Joe
-----Original Message----- From: observium [mailto:observium-bounces@observium.org] On Behalf Of John Brown Sent: Saturday, April 04, 2015 05:56 To: observium Subject: [Observium] observium managment should be more professional
http://afmug.com/pipermail/af/2015-March/018914.html
As a provider of internet connectivity (ergo a telco and a ISP) we use licensed microwave and unlicensed microwave technologies to feed our revenue generating customers. We also use fiber, and copper based technologies.
microwave != wifi
We are not alone in using a mix of technologies to deliver bits to customers. T-Mobile, ATT, Sprint, AllTel, Cellular One, Verizon, etc all use microwave (wireless) technologies to deliver bits.
I, like others, have offered to write code, have asked for pointers on how to get started writing code, all appears to have fallen upon deaf ears.
I've offered to PAY others to write code to support our kit, with features we need. No takers.
As a PAID user, I'm quite shocked and disappointed in Adam's childish behaviour.
Users of Observium should be concerned that their continued investment could be a loosing path as more people move AWAY from the product and go support other packages. On two lists that I saw the above URL, not less than 3 dozen people said they where removing observium and not renewing their annual subscriptions. I suspect that will become a trend.
Cheers _______________________________________________ observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
Adam,
You are terrible at management, create observium WISP edition with 10K per year if they still willing to buy that, you are good to go.
On 05.04.2015 12:10, Adam Armstrong wrote:
All I see in that thread is even more of the same entitled behavior as the idiot who refused to take "no" for an answer for half an hour and got banned from the channel.
This ridiculous drama is based purely on a small section of our user base believing they can demand that we support the devices they use, because, well, they use them. WISPs are a tiny minority of our user base, and an even tinier minority of our paying user base.
There's simply no justification for the huge amount of effort required to build support for these devices to Observium's required standards.
Also, if you read that thread, it's hilariously clear that no one actually read it, because almost no one picks up on the fact that the original poster claims I called all WISPs retarded, when in reality I was referring to the vendors who write the MIBs (which is an undeniable fact to anyone who has ever tried to get information from wireless kit via SNMP)! :DDD
In any event, all this demonstrates to me is that this is a sector of the market which is seemingly way too much hassle to try to support.
adam.
p.s.
Best part of this email is that my mail client marked "cus" as a spelling mistake, since it isn't actually a word, and belies the puritanical scared-of-words origins of the writer.
Not everyone lives in the "must not cuss!" USA!
On 2015-04-05 02:00, John Brown wrote:
Joe, as a person that has written millions of lines of code, starting in the mid-seventies with BASIC, Fortran (both 4 and 77), Pascal, ASM (6502, Z80, 68K, 8080, x86, SPARC, MIPS, etc), C and some C++, along with scripting languages like PHP, Perl, and Python, I can say with reasonable certainty that what is being asked is not more complex or harder than what has already been produced.
Certainly being a jerk, using cus words and generally being highly unprofessional has zip to do with actually writing code. It does have everything to with being an ass.
What the leader of Observium has clearly said is that an entire market segment and technology type is not worth his time and he has done so in a most unprofessional manner. There have been MANY people, myself included, that have been willing to compensate by way of money, spend our time helping develop code, etc, all to be blown off and cus'ed at.
Maybe LibreNMS will be more successful since its a community project, of people, by people and for the community.
Hundreds of people have asked for the ability to add the ability to graph, alert, etc on a OID. Not an unreasonable request. A custom sensor or similar could be created. The customer base has made requests, the keeper of the keys has told them to pound sand.
On Sat, Apr 4, 2015 at 5:27 PM, Joseph L. Brunner joe@affirmedsystems.com wrote:
John,
When you consider how complex it is to "write software" its not something you can take up gradually. A Google fellow ALL OF 32, estimates in his lifetime he has spent 50,000 hours writing software. I'm guessing that is 20 years, perhaps 60 hours a week (since he was a kid), as a "year" only has 8,760 hours.
And even Google the other day had massive gmail outage that CNBC reported during their telecast despite 1000's of pro's like him on staff tasked with their "software".
It's not some nothing you take up in your spare time and have much luck at it, imho.
If the cursory graphs and checks (which are pretty nice) in Observium are not enough, I recommend brining in IBM netcool or HP Openview teams like many major ISP's (Level3, ATT) have had to do. Of course, these consultants rate's start at $250/hour in my experience and not $100 a year like Observium :)
In our similar operation to your own, ran on behalf of private low latency trading firms between NY and CHICAGO, with extremely secretive, precise requirements - we monitor the "overall" network with Observium to get "a sense" of utilization at 20,000 feet - then drill down with very very precise tools written for each purpose in python, java, bash shells, etc - each with its own backend data storage and reporting scheme from a simple as rrd files, log files, pcap files to as complex as Microsoft sql and couchbase.
In short, I think you are trying to do too much with one tool. You need a "polling style" app for the overview time series, and you need "real-time sensors" that trip and alert in a second or less. We use many variants of both. I call it "monitoring in depth". So even if 3 out of 4 of our tools fail - the 4th one will alert us and the client, often with seconds to "go flat". Imagine you were writing call options in microseconds and you didn't have the current stock price or news about the stock :)
Thanks,
Joe
-----Original Message----- From: observium [mailto:observium-bounces@observium.org] On Behalf Of John Brown Sent: Saturday, April 04, 2015 05:56 To: observium Subject: [Observium] observium managment should be more professional
http://afmug.com/pipermail/af/2015-March/018914.html
As a provider of internet connectivity (ergo a telco and a ISP) we use licensed microwave and unlicensed microwave technologies to feed our revenue generating customers. We also use fiber, and copper based technologies.
microwave != wifi
We are not alone in using a mix of technologies to deliver bits to customers. T-Mobile, ATT, Sprint, AllTel, Cellular One, Verizon, etc all use microwave (wireless) technologies to deliver bits.
I, like others, have offered to write code, have asked for pointers on how to get started writing code, all appears to have fallen upon deaf ears.
I've offered to PAY others to write code to support our kit, with features we need. No takers.
As a PAID user, I'm quite shocked and disappointed in Adam's childish behaviour.
Users of Observium should be concerned that their continued investment could be a loosing path as more people move AWAY from the product and go support other packages. On two lists that I saw the above URL, not less than 3 dozen people said they where removing observium and not renewing their annual subscriptions. I suspect that will become a trend.
Cheers _______________________________________________ observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
On 2015-04-05 10:55, Nikolay Shopik wrote:
Adam,
You are terrible at management, create observium WISP edition with 10K per year if they still willing to buy that, you are good to go.
Y'see, the thing is, 10k from a dozen users probably wouldn't cover the effort required to build it!
£150 from a dozen users barely even covers reading through one vendor's MIB!
I think I've spent about 3-4 aggregate days worth of time on reading through MIBs and trying to come up with workable database schemas to accommodate all of the devices.
Wireless kit seems to have been developed by people who had never encountered SNMP, MIBs or ASN.1 before. The MIBs are some of the worst I've seen anywhere, far worse than what you get from the "budget" switch/router vendors. Every wifi vendor has their own idea of what goes with which set of entities and more than one vendor thinks it's a good idea to write a new MIB for every single model they release.
That's assuming they actually output more than a broken IF-MIB, of course.
At the end of the day, these people have no idea of the scale of what they're /demanding/ nor seem to have any understanding of the tiny percentage of the Observium userbase they represent. Observium is primarily a platform written by an ISP/Telco engineer to monitor ISP/Telco kit. These people are acting like a skateboarder who's turned up to a BMX track and is arguing all of the jumps need to be concreted so that he can use them.
For example, the devices that the guy I was arguing on IRC claimed to be super important to support? We have something like 10 of them in our entire userbase. It's just insane, they have no sense of scale.
adam.
Adam Your comments make good Sunday morning reading .
Regards
Watson
Sent from my iPhone
On 05 Apr 2015, at 12:05 PM, Adam Armstrong adama@memetic.org wrote:
On 2015-04-05 10:55, Nikolay Shopik wrote: Adam, You are terrible at management, create observium WISP edition with 10K per year if they still willing to buy that, you are good to go.
Y'see, the thing is, 10k from a dozen users probably wouldn't cover the effort required to build it!
£150 from a dozen users barely even covers reading through one vendor's MIB!
I think I've spent about 3-4 aggregate days worth of time on reading through MIBs and trying to come up with workable database schemas to accommodate all of the devices.
Wireless kit seems to have been developed by people who had never encountered SNMP, MIBs or ASN.1 before. The MIBs are some of the worst I've seen anywhere, far worse than what you get from the "budget" switch/router vendors. Every wifi vendor has their own idea of what goes with which set of entities and more than one vendor thinks it's a good idea to write a new MIB for every single model they release.
That's assuming they actually output more than a broken IF-MIB, of course.
At the end of the day, these people have no idea of the scale of what they're /demanding/ nor seem to have any understanding of the tiny percentage of the Observium userbase they represent. Observium is primarily a platform written by an ISP/Telco engineer to monitor ISP/Telco kit. These people are acting like a skateboarder who's turned up to a BMX track and is arguing all of the jumps need to be concreted so that he can use them.
For example, the devices that the guy I was arguing on IRC claimed to be super important to support? We have something like 10 of them in our entire userbase. It's just insane, they have no sense of scale.
adam. _______________________________________________ observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
Most of the wireless guys are ISPs, however.
----- Mike Hammett Intelligent Computing Solutions http://www.ics-il.com
----- Original Message -----
From: "Adam Armstrong" adama@memetic.org To: "Observium Network Observation System" observium@observium.org Sent: Sunday, April 5, 2015 5:04:58 AM Subject: Re: [Observium] observium managment should be more professional
On 2015-04-05 10:55, Nikolay Shopik wrote:
Adam,
You are terrible at management, create observium WISP edition with 10K per year if they still willing to buy that, you are good to go.
Y'see, the thing is, 10k from a dozen users probably wouldn't cover the effort required to build it!
£150 from a dozen users barely even covers reading through one vendor's MIB!
I think I've spent about 3-4 aggregate days worth of time on reading through MIBs and trying to come up with workable database schemas to accommodate all of the devices.
Wireless kit seems to have been developed by people who had never encountered SNMP, MIBs or ASN.1 before. The MIBs are some of the worst I've seen anywhere, far worse than what you get from the "budget" switch/router vendors. Every wifi vendor has their own idea of what goes with which set of entities and more than one vendor thinks it's a good idea to write a new MIB for every single model they release.
That's assuming they actually output more than a broken IF-MIB, of course.
At the end of the day, these people have no idea of the scale of what they're /demanding/ nor seem to have any understanding of the tiny percentage of the Observium userbase they represent. Observium is primarily a platform written by an ISP/Telco engineer to monitor ISP/Telco kit. These people are acting like a skateboarder who's turned up to a BMX track and is arguing all of the jumps need to be concreted so that he can use them.
For example, the devices that the guy I was arguing on IRC claimed to be super important to support? We have something like 10 of them in our entire userbase. It's just insane, they have no sense of scale.
adam. _______________________________________________ observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
Yes!
On 4/5/2015 7:43 AM, Mike Hammett wrote:
Most of the wireless guys are ISPs, however.
Mike Hammett Intelligent Computing Solutions http://www.ics-il.com
https://www.facebook.com/ICSILhttps://plus.google.com/+IntelligentComputingSolutionsDeKalbhttps://www.linkedin.com/company/intelligent-computing-solutionshttps://twitter.com/ICSIL
*From: *"Adam Armstrong" adama@memetic.org *To: *"Observium Network Observation System" observium@observium.org *Sent: *Sunday, April 5, 2015 5:04:58 AM *Subject: *Re: [Observium] observium managment should be more professional
On 2015-04-05 10:55, Nikolay Shopik wrote:
Adam,
You are terrible at management, create observium WISP edition with 10K per year if they still willing to buy that, you are good to go.
Y'see, the thing is, 10k from a dozen users probably wouldn't cover the effort required to build it!
£150 from a dozen users barely even covers reading through one vendor's MIB!
I think I've spent about 3-4 aggregate days worth of time on reading through MIBs and trying to come up with workable database schemas to accommodate all of the devices.
Wireless kit seems to have been developed by people who had never encountered SNMP, MIBs or ASN.1 before. The MIBs are some of the worst I've seen anywhere, far worse than what you get from the "budget" switch/router vendors. Every wifi vendor has their own idea of what goes with which set of entities and more than one vendor thinks it's a good idea to write a new MIB for every single model they release.
That's assuming they actually output more than a broken IF-MIB, of course.
At the end of the day, these people have no idea of the scale of what they're /demanding/ nor seem to have any understanding of the tiny percentage of the Observium userbase they represent. Observium is primarily a platform written by an ISP/Telco engineer to monitor ISP/Telco kit. These people are acting like a skateboarder who's turned up to a BMX track and is arguing all of the jumps need to be concreted so that he can use them.
For example, the devices that the guy I was arguing on IRC claimed to be super important to support? We have something like 10 of them in our entire userbase. It's just insane, they have no sense of scale.
adam. _______________________________________________ observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
"sorry, we only support shipping"
"but cruise liners are ships!"
"..."
adam.
On 2015-04-05 13:52, Keefe John wrote:
Yes!
On 4/5/2015 7:43 AM, Mike Hammett wrote:
Most of the wireless guys are ISPs, however.
Mike Hammett Intelligent Computing Solutions http://www.ics-il.com [2]
[3] [4] [5] [6]
FROM: "Adam Armstrong" adama@memetic.org TO: "Observium Network Observation System" observium@observium.org SENT: Sunday, April 5, 2015 5:04:58 AM SUBJECT: Re: [Observium] observium managment should be more professional
On 2015-04-05 10:55, Nikolay Shopik wrote:
Adam,
You are terrible at management, create observium WISP edition
with 10K
per year if they still willing to buy that, you are good to go.
Y'see, the thing is, 10k from a dozen users probably wouldn't cover the effort required to build it!
£150 from a dozen users barely even covers reading through one vendor's MIB!
I think I've spent about 3-4 aggregate days worth of time on reading through MIBs and trying to come up with workable database schemas to accommodate all of the devices.
Wireless kit seems to have been developed by people who had never encountered SNMP, MIBs or ASN.1 before. The MIBs are some of the worst I've seen anywhere, far worse than what you get from the "budget" switch/router vendors. Every wifi vendor has their own idea of what goes with which set of entities and more than one vendor thinks it's a good idea to write a new MIB for every single model they release.
That's assuming they actually output more than a broken IF-MIB, of course.
At the end of the day, these people have no idea of the scale of what they're /demanding/ nor seem to have any understanding of the tiny percentage of the Observium userbase they represent. Observium is primarily a platform written by an ISP/Telco engineer to monitor ISP/Telco kit. These people are acting like a skateboarder who's turned up to a BMX track and is arguing all of the jumps need to be concreted so that he can use them.
For example, the devices that the guy I was arguing on IRC claimed to be super important to support? We have something like 10 of them in our entire userbase. It's just insane, they have no sense of scale.
adam. _______________________________________________ observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium [1]
observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium [1]
Links:
[1] http://postman.memetic.org/cgi-bin/mailman/listinfo/observium [2] http://www.ics-il.com [3] https://www.facebook.com/ICSIL [4] https://plus.google.com/+IntelligentComputingSolutionsDeKalb [5] https://www.linkedin.com/company/intelligent-computing-solutions [6] https://twitter.com/ICSIL
observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
You know what I mean.
adam.
On 2015-04-05 13:43, Mike Hammett wrote:
Most of the wireless guys are ISPs, however.
Mike Hammett Intelligent Computing Solutions http://www.ics-il.com
[1] [2] [3] [4]
FROM: "Adam Armstrong" adama@memetic.org TO: "Observium Network Observation System" observium@observium.org SENT: Sunday, April 5, 2015 5:04:58 AM SUBJECT: Re: [Observium] observium managment should be more professional
On 2015-04-05 10:55, Nikolay Shopik wrote:
Adam,
You are terrible at management, create observium WISP edition with
10K
per year if they still willing to buy that, you are good to go.
Y'see, the thing is, 10k from a dozen users probably wouldn't cover the effort required to build it!
£150 from a dozen users barely even covers reading through one vendor's MIB!
I think I've spent about 3-4 aggregate days worth of time on reading through MIBs and trying to come up with workable database schemas to accommodate all of the devices.
Wireless kit seems to have been developed by people who had never encountered SNMP, MIBs or ASN.1 before. The MIBs are some of the worst
I've seen anywhere, far worse than what you get from the "budget" switch/router vendors. Every wifi vendor has their own idea of what goes with which set of entities and more than one vendor thinks it's a good
idea to write a new MIB for every single model they release.
That's assuming they actually output more than a broken IF-MIB, of course.
At the end of the day, these people have no idea of the scale of what they're /demanding/ nor seem to have any understanding of the tiny percentage of the Observium userbase they represent. Observium is primarily a platform written by an ISP/Telco engineer to monitor ISP/Telco kit. These people are acting like a skateboarder who's turned up to a BMX track and is arguing all of the jumps need to be concreted
so that he can use them.
For example, the devices that the guy I was arguing on IRC claimed to be super important to support? We have something like 10 of them in our entire userbase. It's just insane, they have no sense of scale.
adam. _______________________________________________ observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
Links:
[1] https://www.facebook.com/ICSIL [2] https://plus.google.com/+IntelligentComputingSolutionsDeKalb [3] https://www.linkedin.com/company/intelligent-computing-solutions [4] https://twitter.com/ICSIL
observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
You also get no argument from me (and many other wireless ISPs) that the vendors are absolutely terrible at SNMP. When I was working to get a Dragonwave link into Mikrotik's "The Dude" monitoring program, just finding the correct MIBs and their dependencies was a bitch. Ubiquiti AirMax series didn't have a MIB until version 5.6, which is still in beta. Their AirFiber series didn't even have SNMP for quite some time.
*sigh*
We are pretty good at getting what we want out of our vendors, however. Can you put together a page on Observium's web site that is mostly neutral in tone (blatantly calling them fucking idiots to their face might not get us far) regarding generally how these vendors are fucking up and what they should be doing? Then we can expend our energies to get them to fix things.
Keep in mind that there is a world of attributes that wireless operators would want to track that have no parallel in the wired or even WiFi worlds.
----- Mike Hammett Intelligent Computing Solutions http://www.ics-il.com
----- Original Message -----
From: "Adam Armstrong" adama@memetic.org To: "Observium Network Observation System" observium@observium.org Sent: Sunday, April 5, 2015 5:04:58 AM Subject: Re: [Observium] observium managment should be more professional
On 2015-04-05 10:55, Nikolay Shopik wrote:
Adam,
You are terrible at management, create observium WISP edition with 10K per year if they still willing to buy that, you are good to go.
Y'see, the thing is, 10k from a dozen users probably wouldn't cover the effort required to build it!
£150 from a dozen users barely even covers reading through one vendor's MIB!
I think I've spent about 3-4 aggregate days worth of time on reading through MIBs and trying to come up with workable database schemas to accommodate all of the devices.
Wireless kit seems to have been developed by people who had never encountered SNMP, MIBs or ASN.1 before. The MIBs are some of the worst I've seen anywhere, far worse than what you get from the "budget" switch/router vendors. Every wifi vendor has their own idea of what goes with which set of entities and more than one vendor thinks it's a good idea to write a new MIB for every single model they release.
That's assuming they actually output more than a broken IF-MIB, of course.
At the end of the day, these people have no idea of the scale of what they're /demanding/ nor seem to have any understanding of the tiny percentage of the Observium userbase they represent. Observium is primarily a platform written by an ISP/Telco engineer to monitor ISP/Telco kit. These people are acting like a skateboarder who's turned up to a BMX track and is arguing all of the jumps need to be concreted so that he can use them.
For example, the devices that the guy I was arguing on IRC claimed to be super important to support? We have something like 10 of them in our entire userbase. It's just insane, they have no sense of scale.
adam. _______________________________________________ observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
All they really need to do is be consistent, and coming up with a standard wouldn't hurt.
Why it's necessary for every vendor to invent their own special way of putting radios/lans/clients, etc into SNMP, I don't know.
There's also the strange thing most vendors do where data is duplicated quite a lot. They'll duplicate half of the radio data in the wlan table, for no good reason.
Most of the problems are historical and many of the vendors seem to be heading in the right direction, probably down to a lot of whining from their customer base (which they seem to be especially good at, hah!).
At this point I think I've already wasted quite enough of my time on a clearly undesirable market segment.
The usability difference between Observium and Cacti is inseparable from the order of magnitude more effort required to graph new things in Observium. That's just how it is. The reason we are different to other applications is because we have lots of restrictions and unconventional ways of doing things.
adam.
On 2015-04-05 14:10, Mike Hammett wrote:
You also get no argument from me (and many other wireless ISPs) that the vendors are absolutely terrible at SNMP. When I was working to get a Dragonwave link into Mikrotik's "The Dude" monitoring program, just finding the correct MIBs and their dependencies was a bitch. Ubiquiti AirMax series didn't have a MIB until version 5.6, which is still in beta. Their AirFiber series didn't even have SNMP for quite some time.
*sigh*
We are pretty good at getting what we want out of our vendors, however. Can you put together a page on Observium's web site that is mostly neutral in tone (blatantly calling them fucking idiots to their face might not get us far) regarding generally how these vendors are fucking up and what they should be doing? Then we can expend our energies to get them to fix things.
Keep in mind that there is a world of attributes that wireless operators would want to track that have no parallel in the wired or even WiFi worlds.
Mike Hammett Intelligent Computing Solutions http://www.ics-il.com
[1] [2] [3] [4]
FROM: "Adam Armstrong" adama@memetic.org TO: "Observium Network Observation System" observium@observium.org SENT: Sunday, April 5, 2015 5:04:58 AM SUBJECT: Re: [Observium] observium managment should be more professional
On 2015-04-05 10:55, Nikolay Shopik wrote:
Adam,
You are terrible at management, create observium WISP edition with
10K
per year if they still willing to buy that, you are good to go.
Y'see, the thing is, 10k from a dozen users probably wouldn't cover the effort required to build it!
£150 from a dozen users barely even covers reading through one vendor's MIB!
I think I've spent about 3-4 aggregate days worth of time on reading through MIBs and trying to come up with workable database schemas to accommodate all of the devices.
Wireless kit seems to have been developed by people who had never encountered SNMP, MIBs or ASN.1 before. The MIBs are some of the worst
I've seen anywhere, far worse than what you get from the "budget" switch/router vendors. Every wifi vendor has their own idea of what goes with which set of entities and more than one vendor thinks it's a good
idea to write a new MIB for every single model they release.
That's assuming they actually output more than a broken IF-MIB, of course.
At the end of the day, these people have no idea of the scale of what they're /demanding/ nor seem to have any understanding of the tiny percentage of the Observium userbase they represent. Observium is primarily a platform written by an ISP/Telco engineer to monitor ISP/Telco kit. These people are acting like a skateboarder who's turned up to a BMX track and is arguing all of the jumps need to be concreted
so that he can use them.
For example, the devices that the guy I was arguing on IRC claimed to be super important to support? We have something like 10 of them in our entire userbase. It's just insane, they have no sense of scale.
adam. _______________________________________________ observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
Links:
[1] https://www.facebook.com/ICSIL [2] https://plus.google.com/+IntelligentComputingSolutionsDeKalb [3] https://www.linkedin.com/company/intelligent-computing-solutions [4] https://twitter.com/ICSIL
observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
These people are acting like a skateboarder who's turned up to a BMX track and is arguing all of the jumps need to be concreted so that he can use them.
Best line ever...
I'm in total agreement with you Adam
Use Cisco/Juniper/Arista - or you're doing it wrong!
-----Original Message----- From: observium [mailto:observium-bounces@observium.org] On Behalf Of Adam Armstrong Sent: Sunday, April 05, 2015 06:05 AM To: Observium Network Observation System Subject: Re: [Observium] observium managment should be more professional
On 2015-04-05 10:55, Nikolay Shopik wrote:
Adam,
You are terrible at management, create observium WISP edition with 10K per year if they still willing to buy that, you are good to go.
Y'see, the thing is, 10k from a dozen users probably wouldn't cover the effort required to build it!
£150 from a dozen users barely even covers reading through one vendor's MIB!
I think I've spent about 3-4 aggregate days worth of time on reading through MIBs and trying to come up with workable database schemas to accommodate all of the devices.
Wireless kit seems to have been developed by people who had never encountered SNMP, MIBs or ASN.1 before. The MIBs are some of the worst I've seen anywhere, far worse than what you get from the "budget" switch/router vendors. Every wifi vendor has their own idea of what goes with which set of entities and more than one vendor thinks it's a good idea to write a new MIB for every single model they release.
That's assuming they actually output more than a broken IF-MIB, of course.
At the end of the day, these people have no idea of the scale of what they're /demanding/ nor seem to have any understanding of the tiny percentage of the Observium userbase they represent. Observium is primarily a platform written by an ISP/Telco engineer to monitor ISP/Telco kit. These people are acting like a skateboarder who's turned up to a BMX track and is arguing all of the jumps need to be concreted so that he can use them.
For example, the devices that the guy I was arguing on IRC claimed to be super important to support? We have something like 10 of them in our entire userbase. It's just insane, they have no sense of scale.
adam. _______________________________________________ observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
On Sun, Apr 5, 2015 at 10:23 PM, Joseph L. Brunner joe@affirmedsystems.com wrote:
These people are acting like a skateboarder who's turned up to a BMX track and is arguing all of the jumps need to be concreted so that he can use them.
Best line ever...
I'm in total agreement with you Adam
Use Cisco/Juniper/Arista - or you're doing it wrong!
-----Original Message----- From: observium [mailto:observium-bounces@observium.org] On Behalf Of Adam Armstrong Sent: Sunday, April 05, 2015 06:05 AM To: Observium Network Observation System Subject: Re: [Observium] observium managment should be more professional
On 2015-04-05 10:55, Nikolay Shopik wrote:
Adam,
You are terrible at management, create observium WISP edition with 10K per year if they still willing to buy that, you are good to go.
Y'see, the thing is, 10k from a dozen users probably wouldn't cover the effort required to build it!
£150 from a dozen users barely even covers reading through one vendor's MIB!
I think I've spent about 3-4 aggregate days worth of time on reading through MIBs and trying to come up with workable database schemas to accommodate all of the devices.
Wireless kit seems to have been developed by people who had never encountered SNMP, MIBs or ASN.1 before. The MIBs are some of the worst I've seen anywhere, far worse than what you get from the "budget" switch/router vendors. Every wifi vendor has their own idea of what goes with which set of entities and more than one vendor thinks it's a good idea to write a new MIB for every single model they release.
That's assuming they actually output more than a broken IF-MIB, of course.
At the end of the day, these people have no idea of the scale of what they're /demanding/ nor seem to have any understanding of the tiny percentage of the Observium userbase they represent. Observium is primarily a platform written by an ISP/Telco engineer to monitor ISP/Telco kit. These people are acting like a skateboarder who's turned up to a BMX track and is arguing all of the jumps need to be concreted so that he can use them.
For example, the devices that the guy I was arguing on IRC claimed to be super important to support? We have something like 10 of them in our entire userbase. It's just insane, they have no sense of scale.
So why not, instead of just being an ahole about it, say it directly: what you ask for requires X hours times Y dollars. Every update you will want will be Z dollars on top of that. And let the people demanding the feature make their own calculations on whether or not they want to go ahead and cash out? In my experience building large complex systems, this approach keeps everyone happy, and quite often turns into a well paid for project.
adam. _______________________________________________ observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium _______________________________________________ observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
None of those companies make gear for our space and the wireless stuff they do make gets laughed at.
----- Mike Hammett Intelligent Computing Solutions http://www.ics-il.com
----- Original Message -----
From: "Joseph L. Brunner" joe@affirmedsystems.com To: "Observium Network Observation System" observium@observium.org Sent: Sunday, April 5, 2015 9:23:12 PM Subject: Re: [Observium] observium managment should be more professional
These people are acting like a skateboarder who's turned up to a BMX track and is arguing all of the jumps need to be concreted so that he can use them.
Best line ever...
I'm in total agreement with you Adam
Use Cisco/Juniper/Arista - or you're doing it wrong!
-----Original Message----- From: observium [mailto:observium-bounces@observium.org] On Behalf Of Adam Armstrong Sent: Sunday, April 05, 2015 06:05 AM To: Observium Network Observation System Subject: Re: [Observium] observium managment should be more professional
On 2015-04-05 10:55, Nikolay Shopik wrote:
Adam,
You are terrible at management, create observium WISP edition with 10K per year if they still willing to buy that, you are good to go.
Y'see, the thing is, 10k from a dozen users probably wouldn't cover the effort required to build it!
£150 from a dozen users barely even covers reading through one vendor's MIB!
I think I've spent about 3-4 aggregate days worth of time on reading through MIBs and trying to come up with workable database schemas to accommodate all of the devices.
Wireless kit seems to have been developed by people who had never encountered SNMP, MIBs or ASN.1 before. The MIBs are some of the worst I've seen anywhere, far worse than what you get from the "budget" switch/router vendors. Every wifi vendor has their own idea of what goes with which set of entities and more than one vendor thinks it's a good idea to write a new MIB for every single model they release.
That's assuming they actually output more than a broken IF-MIB, of course.
At the end of the day, these people have no idea of the scale of what they're /demanding/ nor seem to have any understanding of the tiny percentage of the Observium userbase they represent. Observium is primarily a platform written by an ISP/Telco engineer to monitor ISP/Telco kit. These people are acting like a skateboarder who's turned up to a BMX track and is arguing all of the jumps need to be concreted so that he can use them.
For example, the devices that the guy I was arguing on IRC claimed to be super important to support? We have something like 10 of them in our entire userbase. It's just insane, they have no sense of scale.
adam. _______________________________________________ observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium _______________________________________________ observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
I don't agree with the Use C/J/A line per se - the message to be conveyed is this:
Use whatever you want, but if Observium does not support your gear, either add support(*) or don't use Observium.
(*) either by paying someone to do it(**) or by doing it yourselfr (**) you can't force anyone to anything, not even with money - so if noone on the Observium team wants to do it, you'll have to find someone else
I've kept out of this discussion for a while, but it really is that easy. Observium is good (if I say so myself) at what it does - if you want something that it doesn't do, maybe it's not for you. No need to go all whiny on the IRC channel over it, or even more ridiculous, paste half IRC conversations on OTHER mailing lists crying like a baby; the person it reflects most badly on in this instance is NOT Adam.
Tom
On 04/06/2015 04:37 AM, Mike Hammett wrote:
None of those companies make gear for our space and the wireless stuff they do make gets laughed at.
Mike Hammett Intelligent Computing Solutions http://www.ics-il.com
https://www.facebook.com/ICSILhttps://plus.google.com/+IntelligentComputingSolutionsDeKalbhttps://www.linkedin.com/company/intelligent-computing-solutionshttps://twitter.com/ICSIL
*From: *"Joseph L. Brunner" joe@affirmedsystems.com *To: *"Observium Network Observation System" observium@observium.org *Sent: *Sunday, April 5, 2015 9:23:12 PM *Subject: *Re: [Observium] observium managment should be more professional
These people are acting like a skateboarder who's turned up to a BMX track and is arguing all of the jumps need to be concreted so that he can use them.
Best line ever...
I'm in total agreement with you Adam
Use Cisco/Juniper/Arista - or you're doing it wrong!
-----Original Message----- From: observium [mailto:observium-bounces@observium.org] On Behalf Of Adam Armstrong Sent: Sunday, April 05, 2015 06:05 AM To: Observium Network Observation System Subject: Re: [Observium] observium managment should be more professional
On 2015-04-05 10:55, Nikolay Shopik wrote:
Adam,
You are terrible at management, create observium WISP edition with 10K per year if they still willing to buy that, you are good to go.
Y'see, the thing is, 10k from a dozen users probably wouldn't cover the effort required to build it!
£150 from a dozen users barely even covers reading through one vendor's MIB!
I think I've spent about 3-4 aggregate days worth of time on reading through MIBs and trying to come up with workable database schemas to accommodate all of the devices.
Wireless kit seems to have been developed by people who had never encountered SNMP, MIBs or ASN.1 before. The MIBs are some of the worst I've seen anywhere, far worse than what you get from the "budget" switch/router vendors. Every wifi vendor has their own idea of what goes with which set of entities and more than one vendor thinks it's a good idea to write a new MIB for every single model they release.
That's assuming they actually output more than a broken IF-MIB, of course.
At the end of the day, these people have no idea of the scale of what they're /demanding/ nor seem to have any understanding of the tiny percentage of the Observium userbase they represent. Observium is primarily a platform written by an ISP/Telco engineer to monitor ISP/Telco kit. These people are acting like a skateboarder who's turned up to a BMX track and is arguing all of the jumps need to be concreted so that he can use them.
For example, the devices that the guy I was arguing on IRC claimed to be super important to support? We have something like 10 of them in our entire userbase. It's just insane, they have no sense of scale.
adam. _______________________________________________ observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium _______________________________________________ observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
participants (9)
-
Adam Armstrong -
Dan Yasny -
John Brown -
Joseph L. Brunner -
Keefe John -
Mike Hammett -
Nikolay Shopik -
Tom Laermans -
Watson Kamanga