Network Admin in the Himalayas
Location: Nangi, Nepal
Local Time: Friday, Dec 23, 2005 - 9:00am
You know, I have never been a network admin before, so maybe this kind of repetative work is typical, but it seems to me there are particular challenges to keeping the network going up here.
Internet only lasts until a short time after dark, because some of the relays are solar powered with minimal battery power. Yesterday we got the Internet working again after 2 days without. During that time we knew (or at least I knew, people were questioning my judgement, but it was pretty clear to me which relay was out) the problem was in Pokhara. However, without the network connection to Pokhara, we could not contact them to ask them to look at the problem with the network, since the only phone in town runs over the network. So, we didn't even know if they knew about the problem.
It is a full day journey to get to Pokhara from here. At least 4 hours walking (if you go fast), a good 6 hour bus ride. Coming back is worse because the walking trip from Beni (where the bus stops) involves a good 3 hours up a 60 - 70 degree hill.
So, finally the Internet started working again. Pokhara phoned us moments later to tell us that the hub had gone down and they had to replace it.
This morning, I come over here at about 8am. No Internet connection again.
This time I can see there is a problem with our radio. We play with the power connection, and the connection to the switch, and then switch the cable between inside the building and outside the building where the radio is, pointing up the mountain at Relay station 1.
We figure out that, over night, for some reason, the cable from inside the building to outside the building has stopped passing signal. Maybe the mice chewed it, maybe alternating heat and cold (sub zero temperatures here each night) messed it up. Basically I'm just guessing though, because all I know for sure is last night it worked, today it doesn't.
Another situation, Leal was working on one of the IP phones this morning. He kept moving the power plug from one socket to another because for some reason the phone wouldn't power up. Just because a power socket worked last night, is no reason to assume it is working today, so he was trying different power sockets, trying to get the phone to power up.
Finally he noticed that a mouse had chewed about 2 inches of the power cable. Over night. That phone worked yesterday.
It seems that nearly every day, work is required just to keep basic Internet connectivity going. A cable stops working. In almost 7 years of professional programming, this has happened to me once that I know of. Here in the Himalayas, I've seen it several times in less than a month.
A radio turns off. Who knows why? Some times it seems that the voltage going to the radio is not enough so the radio gets in a wierd state where the power light is on, but there is not enough power for it to function. Everything looks okay, there is link light, but no response to a ping. Unplug the radio power for 10 seconds and plug back in to reset it. Bob's your uncle and away you go.
A cable connector gets loose. The connection can be fine one day, and no link the next.
As a programmer each time I see a problem with a running program, I think what can be changed so this doesn't happen again. Each time I spend a day figuring out that some strange message, or even a blank screen, means a line is missing from a config file, I work to make a meaningful error message, so that next time the problem occurs I get an error message on the screen that allows me to immediately fix the problem without spending 6 hours figuring out what the problem is.
However, that is not how it works here. Some thing can be working perfectly, and the next day a mouse chewed a cable, or cold weather affected the equipment, or a power surge at night cooked the hub. Power here is specified to be in the range of 100 - 240V, but it sometimes surges over 300. In only 1 month here, I've seen monitors and power adaptors suddenly start to smoke.
I don't know what you do to fix those problems forever. You just deal with each problem as it comes up. Each day you fix what happened and go on.
One of the reasons I came here to Nangi was I wanted to see a place that was using the Internet the way I think it should be used, for education and communication. If Nangi can do it, why not other places in the world. To me, education is the process of making you aware of choices. You've got to learn to read and write, and some basic math skills, but the details of what they teach you beyond that are mostly irrelevant. It is the process of learning, cracking your mind open, making it change shape, so that you have the ability to see possibilities, to know that things exist beyond the scope of your experience. That is the important thing.
The Internet is part of that. Or it should be. It is access to vast amounts of information. Consider the library resources necessary to research plants appropriate for a nitrogen rich environment, such as a urine bed. Alternatively, about 10 minutes of work with Google can get you some answers.
What I'm finding is that access is not cheap or easy, not as we take it for granted in Canada.
I think there are about a dozen villages sharing this Internet connection. There are few people who work on this network full time, and several others who contribute part time to keep it going every single day.
The link to the Internet is a 64kps link that costs 12,000Rs (~ $240CND) each month. ADSL (Telus High Speed Internet) in Canada gives 640kps in it's slowest direction, for $40CND each month. Most of us don't share that with anyone, certain not with 40 other computers.
The major problem of trying to supply something like Internet access to a remote village is sustainability. There is work to do every day. The village has to want the access, and be willing to put work into it.
Well, that's my ramble for the moment.
Local Time: Friday, Dec 23, 2005 - 9:00am
You know, I have never been a network admin before, so maybe this kind of repetative work is typical, but it seems to me there are particular challenges to keeping the network going up here.
Internet only lasts until a short time after dark, because some of the relays are solar powered with minimal battery power. Yesterday we got the Internet working again after 2 days without. During that time we knew (or at least I knew, people were questioning my judgement, but it was pretty clear to me which relay was out) the problem was in Pokhara. However, without the network connection to Pokhara, we could not contact them to ask them to look at the problem with the network, since the only phone in town runs over the network. So, we didn't even know if they knew about the problem.
It is a full day journey to get to Pokhara from here. At least 4 hours walking (if you go fast), a good 6 hour bus ride. Coming back is worse because the walking trip from Beni (where the bus stops) involves a good 3 hours up a 60 - 70 degree hill.
So, finally the Internet started working again. Pokhara phoned us moments later to tell us that the hub had gone down and they had to replace it.
This morning, I come over here at about 8am. No Internet connection again.
This time I can see there is a problem with our radio. We play with the power connection, and the connection to the switch, and then switch the cable between inside the building and outside the building where the radio is, pointing up the mountain at Relay station 1.
We figure out that, over night, for some reason, the cable from inside the building to outside the building has stopped passing signal. Maybe the mice chewed it, maybe alternating heat and cold (sub zero temperatures here each night) messed it up. Basically I'm just guessing though, because all I know for sure is last night it worked, today it doesn't.
Another situation, Leal was working on one of the IP phones this morning. He kept moving the power plug from one socket to another because for some reason the phone wouldn't power up. Just because a power socket worked last night, is no reason to assume it is working today, so he was trying different power sockets, trying to get the phone to power up.
Finally he noticed that a mouse had chewed about 2 inches of the power cable. Over night. That phone worked yesterday.
It seems that nearly every day, work is required just to keep basic Internet connectivity going. A cable stops working. In almost 7 years of professional programming, this has happened to me once that I know of. Here in the Himalayas, I've seen it several times in less than a month.
A radio turns off. Who knows why? Some times it seems that the voltage going to the radio is not enough so the radio gets in a wierd state where the power light is on, but there is not enough power for it to function. Everything looks okay, there is link light, but no response to a ping. Unplug the radio power for 10 seconds and plug back in to reset it. Bob's your uncle and away you go.
A cable connector gets loose. The connection can be fine one day, and no link the next.
As a programmer each time I see a problem with a running program, I think what can be changed so this doesn't happen again. Each time I spend a day figuring out that some strange message, or even a blank screen, means a line is missing from a config file, I work to make a meaningful error message, so that next time the problem occurs I get an error message on the screen that allows me to immediately fix the problem without spending 6 hours figuring out what the problem is.
However, that is not how it works here. Some thing can be working perfectly, and the next day a mouse chewed a cable, or cold weather affected the equipment, or a power surge at night cooked the hub. Power here is specified to be in the range of 100 - 240V, but it sometimes surges over 300. In only 1 month here, I've seen monitors and power adaptors suddenly start to smoke.
I don't know what you do to fix those problems forever. You just deal with each problem as it comes up. Each day you fix what happened and go on.
One of the reasons I came here to Nangi was I wanted to see a place that was using the Internet the way I think it should be used, for education and communication. If Nangi can do it, why not other places in the world. To me, education is the process of making you aware of choices. You've got to learn to read and write, and some basic math skills, but the details of what they teach you beyond that are mostly irrelevant. It is the process of learning, cracking your mind open, making it change shape, so that you have the ability to see possibilities, to know that things exist beyond the scope of your experience. That is the important thing.
The Internet is part of that. Or it should be. It is access to vast amounts of information. Consider the library resources necessary to research plants appropriate for a nitrogen rich environment, such as a urine bed. Alternatively, about 10 minutes of work with Google can get you some answers.
What I'm finding is that access is not cheap or easy, not as we take it for granted in Canada.
I think there are about a dozen villages sharing this Internet connection. There are few people who work on this network full time, and several others who contribute part time to keep it going every single day.
The link to the Internet is a 64kps link that costs 12,000Rs (~ $240CND) each month. ADSL (Telus High Speed Internet) in Canada gives 640kps in it's slowest direction, for $40CND each month. Most of us don't share that with anyone, certain not with 40 other computers.
The major problem of trying to supply something like Internet access to a remote village is sustainability. There is work to do every day. The village has to want the access, and be willing to put work into it.
Well, that's my ramble for the moment.
posted by Jaye | 8:39 AM
0 Comments:
Post a Comment
<< Home