Trip log

14 views
Skip to first unread message

Asheesh Sharma

unread,
Oct 29, 2013, 6:07:21 AM10/29/13
to ruralne...@googlegroups.com
Sir,
    Please find the attachment 'log.rtf'.

Regards,
Asheesh Sharma
log.rtf

Asheesh Sharma

unread,
Oct 29, 2013, 6:19:05 AM10/29/13
to ruralne...@googlegroups.com
Sir,
Please ignore the above attachment. Please find the attachment below

Regards,
Asheesh
log.rtf

Aaditeshwar Seth

unread,
Oct 29, 2013, 12:53:26 PM10/29/13
to ruralne...@googlegroups.com
Thanks Asheeh, it sounds really adventurous! And I hope you took some
pictures of the locations where the nodes were placed? Please post all
pictures too.

We know that the current setup does not give any incentives to the
PRADAN staff to manage the system, therefore in the long run we should
move towards a system of having the Netbook act as an Internet access
point that can be used by the staff.

Other things from your notes:
- web10g installation: After your fixes, do I assume that we now know
how to do it easily?
- tcptraceroute data not being recorded: Is this fixed? The code should
have gone out after rigorous testing
- Modem problem: Zahir, can you take a look. I don't understand the
details here

Can you send a report as of now, of which nodes are working with which
providers. And the transition plan, of where to move the nodes or
install new nodes.

Arvind: Has data started coming from the new nodes you installed?

Rahul, Sameer: Are we now analyzing data from the new nodes.

All: Please send a revised activity plan with timelines.
- node deployment
- traceroute data
- CDN data
- scripts for data analysis checked in gerrit with a README: throughput,
availability, etc
- web10g analysis of long flow breaks
- buffer behavior tests
- web10g analysis of TCP download and upload
- scheduling policies at gateways
- one time tests
- wrap up, to move towards Android and Windows version of the
measurement suite, and enabling the Netbook as an Internet access point

Aaditeshwar
> --
> You received this message because you are subscribed to the Google
> Groups "ruralnet-act4d" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to ruralnet-act4...@googlegroups.com.
> For more options, visit https://groups.google.com/groups/opt_out.

--
Aaditeshwar Seth
Co-founder, Gram Vaani Community Media

http://gramvaani.org
http://www.linkedin.com/company/gram-vaani-community-media

Arvind Mahla

unread,
Oct 30, 2013, 2:35:02 PM10/30/13
to ruralne...@googlegroups.com
Sir,

Data is there from the Hanumanpura village node. Airtel is performing consistently, but Idea dongle becomes unresponsive frequently and needs  unplugging and replugging.

Web10g Data Analysis:
The web10g data is there for around 4-5 days and I am going through the tcpdumps for finding long flow gaps and eventually analyse them using web10g logs.

Node Deployment:
The date of my travel to install Jaipur & Sikar nodes is yet not finalized but in all probability it will be within 7-8 days.

Will post the timelines for other tasks after discussion with Sameer, Rahul, Asheesh.

Arvind



To unsubscribe from this group and stop receiving emails from it, send an email to ruralnet-act4d+unsubscribe@googlegroups.com.

For more options, visit https://groups.google.com/groups/opt_out.
--
Aaditeshwar Seth
Co-founder, Gram Vaani Community Media

http://gramvaani.org
http://www.linkedin.com/company/gram-vaani-community-media
--
You received this message because you are subscribed to the Google Groups "ruralnet-act4d" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ruralnet-act4d+unsubscribe@googlegroups.com.

Asheesh Sharma

unread,
Oct 30, 2013, 3:20:19 PM10/30/13
to ruralne...@googlegroups.com

Sir,


It was sure adventurous, here are some fun facts:

1. There is a 22 km radius forest, which is not under any protection. This tribe called 'Daund' live there. I got a chance to meet them, they are really humble, they offered a white-whisky like liquor. They are on a protest to stop some factories in the eastern region of the jungle which is affecting Narmada. 

2. Nainpur is the biggest narrow gauge railway station across Asia.


But i had a poor camera, and i don't try unless its a SLR. :/


Web10g: Yes we have better solution. I am working on a standalone package.


tcptraceroute: Yes, it was fixed.


Modem problem: Zahir sir, i have those dongles. Please collect them and troubleshoot.


Lamta, Paraswada, Smanapur, Amarpur and Dindori, all nodes are up. Except for Samnapur's reliance wasn't replaced, as it has no coverage in the area. 


I have a suggestion regarding incentive plan for PRADAN. I talked to Chaitanya @Paraswada, about how we can help in their initiatives. He suggested building a Weather app. So i think we can develop a program which reports of bad weather by using donlge's capability by SMS. Furthermore, a PRADAN executive can create messages without touching the netbook, by SMSing to the dongle, which is then distributed to the mass.


Asheesh

Aaditeshwar Seth

unread,
Oct 30, 2013, 10:17:45 PM10/30/13
to ruralne...@googlegroups.com
Your thoughts on incentives are excellent. And I think we can do this very easily. Zahir, what do you think? We may get away with just installing GRINS on the Netbook. And help them start some programs such as send weather updates to their farmer groups, run a voice helpline, etc.

But please take pictures. We should systematically document the learning from these visits and the kind of problems we encounter.

All: I'll expect to see processed data from all these locations then -- Lamta, Paraswada, Smanapur, Amarpur, Dindori, and Delhi. Plus, the new nodes at Mandla and Hanumapura. And I think Zahir installed a node at Ahmedabad? By processed data I mean data reflected in reports for traceroute, CDN, availability, web10g, buffer behavior. All data to be produced by analysis scripts that automatically spit out graphs and feed into the website. Please send me revised timelines for the various activities I listed.

Aaditeshwar
--
You received this message because you are subscribed to the Google Groups "ruralnet-act4d" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ruralnet-act4...@googlegroups.com.

For more options, visit https://groups.google.com/groups/opt_out.

SAMEER A PANDIT

unread,
Nov 4, 2013, 3:35:57 PM11/4/13
to ruralne...@googlegroups.com
Sir,
Please find the revised timelines inline.

On 29.10.2013 22:23, Aaditeshwar Seth wrote:
> Thanks Asheeh, it sounds really adventurous! And I hope you took some
> pictures of the locations where the nodes were placed? Please post all
> pictures too.
>
> We know that the current setup does not give any incentives to the
> PRADAN staff to manage the system, therefore in the long run we should
> move towards a system of having the Netbook act as an Internet access
> point that can be used by the staff.
>
> Other things from your notes:
> - web10g installation: After your fixes, do I assume that we now know
> how to do it easily?
> - tcptraceroute data not being recorded: Is this fixed? The code
> should have gone out after rigorous testing
> - Modem problem: Zahir, can you take a look. I don't understand the
> details here
>
> Can you send a report as of now, of which nodes are working with which
> providers. And the transition plan, of where to move the nodes or
> install new nodes.
>
> Arvind: Has data started coming from the new nodes you installed?
>
> Rahul, Sameer: Are we now analyzing data from the new nodes.
>
> All: Please send a revised activity plan with timelines.
> - node deployment
New node for Hanumanpura (already installed with airtel,idea
working),Jaipur (handed over to Arvind)

> - traceroute data
12 Nov 2013
Owner:Sameer
> - CDN data
15 Nov 2013
Owner : Sameer

> - scripts for data analysis checked in git with a README:
> throughput, availability, etc
15 Nov 2013
Owner: Rahul and Sameer

> - web10g analysis of long flow breaks
> - buffer behavior tests

> - Buffer size Analysis
15 Nov 2013
Owner: Rahul

> - web10g analysis of TCP download and upload
> - scheduling policies at gateways

> - one time tests
One time test suite: 8 Nov
Owner: Sameer

> - wrap up, to move towards Android and Windows version of the
> measurement suite, and enabling the Netbook as an Internet access
> point
>
> Aaditeshwar
>
> On 29-10-2013 15:49, Asheesh Sharma wrote:
>> Sir,
>> Please ignore the above attachment. Please find the attachment below
>>
>> Regards,
>> Asheesh
>> -- You received this message because you are subscribed to the Google
>> Groups "ruralnet-act4d" group.
>> To unsubscribe from this group and stop receiving emails from it, send
>> an email to ruralnet-act4...@googlegroups.com.
>> For more options, visit https://groups.google.com/groups/opt_out.
>
> --
> Aaditeshwar Seth
> Co-founder, Gram Vaani Community Media
>
> http://gramvaani.org
> http://www.linkedin.com/company/gram-vaani-community-media

--
SAMEER A PANDIT
Mtech 2nd Year
2012MCS2821

Aaditeshwar Seth

unread,
Nov 10, 2013, 11:11:53 AM11/10/13
to ruralne...@googlegroups.com
Can the others reply with their timelines as well.

What is the status of operating nodes? Can we have the daily emails
coming as well.

Arvind Mahla

unread,
Nov 11, 2013, 12:40:58 AM11/11/13
to ruralne...@googlegroups.com
Hi,

Sikar & Jaipur Nodes have been placed and started on 9th Nov evening.

Arvind



To unsubscribe from this group and stop receiving emails from it, send an email to ruralnet-act4d+unsubscribe@googlegroups.com.

For more options, visit https://groups.google.com/groups/opt_out.

--
Aaditeshwar Seth
Co-founder, Gram Vaani Community Media

http://gramvaani.org
http://www.linkedin.com/company/gram-vaani-community-media

--
Aaditeshwar Seth
Co-founder, Gram Vaani Community Media

http://gramvaani.org
http://www.linkedin.com/company/gram-vaani-community-media

--
You received this message because you are subscribed to the Google Groups "ruralnet-act4d" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ruralnet-act4d+unsubscribe@googlegroups.com.

Aaditeshwar Seth

unread,
Nov 11, 2013, 1:20:25 AM11/11/13
to ruralne...@googlegroups.com
Please update on other deliverables as well
To unsubscribe from this group and stop receiving emails from it, send an email to ruralnet-act4...@googlegroups.com.

For more options, visit https://groups.google.com/groups/opt_out.

--
Aaditeshwar Seth
Co-founder, Gram Vaani Community Media

http://gramvaani.org
http://www.linkedin.com/company/gram-vaani-community-media

--
Aaditeshwar Seth
Co-founder, Gram Vaani Community Media

http://gramvaani.org
http://www.linkedin.com/company/gram-vaani-community-media

--
You received this message because you are subscribed to the Google Groups "ruralnet-act4d" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ruralnet-act4...@googlegroups.com.

For more options, visit https://groups.google.com/groups/opt_out.

--
You received this message because you are subscribed to the Google Groups "ruralnet-act4d" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ruralnet-act4...@googlegroups.com.

For more options, visit https://groups.google.com/groups/opt_out.

Aaditeshwar Seth

unread,
Nov 17, 2013, 10:29:18 AM11/17/13
to ruralne...@googlegroups.com
Folks, Please update on your deliverables. Most of the list stands pending. If you are facing problems then please ask.

Rahul Nishant

unread,
Nov 18, 2013, 1:59:03 PM11/18/13
to ruralne...@googlegroups.com
Hi Sir,
Please find the updates inline. The graphs and data have been attached
for availability and Buffersize.
We are working on the update mails and we will try to start in two more
days.

Timeline for update mails : 20 Nov, 2013
Owners : Rahul & Sameer

Thanks & regards,
Rahul Nishant
-The Analysis Scripts has been commited to the git
(nets...@10.208.26.25:/home/netspeed/Repository/ruralnet.git)
-The documentation for these scripts have been updated
- Have attached the current Availability data(Availability.png &
AvailabilityGraph.png) and here are some of the observations
a) The previous availability analysis done for the paper is replicated
b) The data from Jaipur is comparable to the data of the semi-urban
location ,i.e.,Dindori for Idea and MTNL. Idea performs better than
others here also.
c) The data from Hanumanpura shows low connectivity. But, we need more
data to comment on in a better way

>> - web10g analysis of long flow breaks
>> - buffer behavior tests
>>
>> - Buffer size Analysis
>> 15 Nov 2013
>> Owner: Rahul
Have attached the graphs for buffer analysis. These are some of the
observations are as follows:
a-> Downlink Buffer
- Airtel has comparitively high buffersize than others in the rural
areas
- Idea has less buffer size compared to others
- Urban location of Delhi has Bigger buffers than the rural locations

b-> Uplink Buffer
- The uplink buffer sizes are quite similar for all the operators
- Reliance has less uplink buffer than others in Delhi
- Uplink MTNL buffers are less than others in most locations

>> - web10g analysis of TCP download and upload
>> - scheduling policies at gateways
>>
>> - one time tests
>> One time test suite: 8 Nov
>> Owner: Sameer
>>
>> - wrap up, to move towards Android and Windows version of the
>> measurement suite, and enabling the Netbook as an Internet access
>> point
>>
>> Aaditeshwar
>>
>> On 29-10-2013 15:49, Asheesh Sharma wrote:
>> Sir,
>> Please ignore the above attachment. Please find the attachment
>> below
>>
>> Regards,
>> Asheesh
>> -- You received this message because you are subscribed to the
>> Google Groups "ruralnet-act4d" group.
>> To unsubscribe from this group and stop receiving emails from it,
>> send an email to ruralnet-act4...@googlegroups.com.
>> For more options, visit https://groups.google.com/groups/opt_out
>> [1].
>>
>> --
>> Aaditeshwar Seth
>> Co-founder, Gram Vaani Community Media
>>
>> http://gramvaani.org [2]
>> http://www.linkedin.com/company/gram-vaani-community-media [3]
>
> --
> Aaditeshwar Seth
> Co-founder, Gram Vaani Community Media
>
> http://gramvaani.org [2]
> http://www.linkedin.com/company/gram-vaani-community-media [3]
>
> --
> You received this message because you are subscribed to the Google
> Groups "ruralnet-act4d" group.
> To unsubscribe from this group and stop receiving emails from it,
> send an email to ruralnet-act4...@googlegroups.com.
> For more options, visit https://groups.google.com/groups/opt_out [1].
>
> --
> You received this message because you are subscribed to the Google
> Groups "ruralnet-act4d" group.
> To unsubscribe from this group and stop receiving emails from it,
> send an email to ruralnet-act4...@googlegroups.com.
> For more options, visit https://groups.google.com/groups/opt_out [1].
>
> --
> Aaditeshwar Seth
> Co-founder, Gram Vaani Community Media
>
> http://gramvaani.org [2]
> http://www.linkedin.com/company/gram-vaani-community-media [3]
> --
> You received this message because you are subscribed to the Google
> Groups "ruralnet-act4d" group.
> To unsubscribe from this group and stop receiving emails from it,
> send an email to ruralnet-act4...@googlegroups.com.
> For more options, visit https://groups.google.com/groups/opt_out [1].
>
> --
> Aaditeshwar Seth
> Co-founder, Gram Vaani Community Media
>
> http://gramvaani.org [2]
> http://www.linkedin.com/company/gram-vaani-community-media [3]
>
> --
> You received this message because you are subscribed to the Google
> Groups "ruralnet-act4d" group.
> To unsubscribe from this group and stop receiving emails from it,
> send an email to ruralnet-act4...@googlegroups.com.
> For more options, visit https://groups.google.com/groups/opt_out [1].
>
>
> Links:
> ------
> [1] https://groups.google.com/groups/opt_out
> [2] http://gramvaani.org
> [3] http://www.linkedin.com/company/gram-vaani-community-media
AvailabilityGraph.png
DownlinkData.png
Uplink.png
UplinkData.png
Availabilitydata.png
AvailabilityGraph.png

Aaditeshwar Seth

unread,
Nov 18, 2013, 9:03:01 PM11/18/13
to ruralne...@googlegroups.com
1. Availability: Can we have comparison data for ALL operators with
their service in Delhi, Ahmedabad, and Jaipur. This is specifically to
contrast whether rural locations have less availability than urban

2. Buffer behavior: There are more parameters that you need to show
here. The pattern that seemed to be occurring is that initially the
buffer would get full, and then allow more data in smaller quanta. You
have reported the capacity but not the refill quanta and frequency.

3. Traceroute: We were waiting for traceroute data from more locations
and operators I think, to be able to make some concrete claims

4. Throughput: Are all your scripts automated now to produce graphs such
as the ones in the paper?

5. CDN data analysis, pending

6. web10g data analysis of long flow breaks, pending

7. web10g data analysis of TCP to compare with buffer behavior noticed
in the UDP tests, pending

8. Scheduling policy at gateways

9. Modem problem reported by Asheesh, what is the status now?

10. One time test suite
http://www.linkedin.com/company/gram-vaani-community-media

Arvind Mahla

unread,
Nov 18, 2013, 10:19:35 PM11/18/13
to ruralne...@googlegroups.com
Sir,

Please excuse for the delay from my side. I am working on and will finish the 6,7 & 8 tasks on priority and report the results as they are available by 22nd Nov.

Task 6 was deliberately delayed to have good amount of data before going into searching and analysing long flow gaps. 

Generated data for Task 8. Will analyse and report.

Arvind




For more options, visit https://groups.google.com/groups/opt_out
[1].

--
Aaditeshwar Seth
Co-founder, Gram Vaani Community Media

http://gramvaani.org [2]
http://www.linkedin.com/company/gram-vaani-community-media [3]

 --
 Aaditeshwar Seth
 Co-founder, Gram Vaani Community Media

 http://gramvaani.org [2]
 http://www.linkedin.com/company/gram-vaani-community-media [3]

 --
 You received this message because you are subscribed to the Google
Groups "ruralnet-act4d" group.
 To unsubscribe from this group and stop receiving emails from it,

 For more options, visit https://groups.google.com/groups/opt_out [1].

 --
 You received this message because you are subscribed to the Google
Groups "ruralnet-act4d" group.
 To unsubscribe from this group and stop receiving emails from it,

 For more options, visit https://groups.google.com/groups/opt_out [1].

--
Aaditeshwar Seth
Co-founder, Gram Vaani Community Media

http://gramvaani.org [2]
http://www.linkedin.com/company/gram-vaani-community-media [3]
 --
 You received this message because you are subscribed to the Google
Groups "ruralnet-act4d" group.
 To unsubscribe from this group and stop receiving emails from it,

 For more options, visit https://groups.google.com/groups/opt_out [1].

--
Aaditeshwar Seth
Co-founder, Gram Vaani Community Media

http://gramvaani.org [2]
http://www.linkedin.com/company/gram-vaani-community-media [3]

 --
 You received this message because you are subscribed to the Google
Groups "ruralnet-act4d" group.
 To unsubscribe from this group and stop receiving emails from it,

--
Aaditeshwar Seth
Co-founder, Gram Vaani Community Media

http://gramvaani.org
http://www.linkedin.com/company/gram-vaani-community-media

--
You received this message because you are subscribed to the Google Groups "ruralnet-act4d" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ruralnet-act4d+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

SAMEER A PANDIT

unread,
Nov 20, 2013, 9:46:37 PM11/20/13
to ruralne...@googlegroups.com
Sir,
Please find the updates inline and attached report for Traceroute
analysis.

On 19.11.2013 07:33, Aaditeshwar Seth wrote:
> 1. Availability: Can we have comparison data for ALL operators with
> their service in Delhi, Ahmedabad, and Jaipur. This is specifically to
> contrast whether rural locations have less availability than urban
>
> 2. Buffer behavior: There are more parameters that you need to show
> here. The pattern that seemed to be occurring is that initially the
> buffer would get full, and then allow more data in smaller quanta. You
> have reported the capacity but not the refill quanta and frequency.
>
> 3. Traceroute: We were waiting for traceroute data from more locations
> and operators I think, to be able to make some concrete claims
Attached report for Traceroute route length and route last hop latency.
>
> 4. Throughput: Are all your scripts automated now to produce graphs
> such as the ones in the paper?
>
> 5. CDN data analysis, pending
Will be sending the report by today night.
report21Nov.pdf

Aaditeshwar Seth

unread,
Nov 21, 2013, 1:12:37 AM11/21/13
to ruralne...@googlegroups.com
I think we were also discussing things like the number of exit gateways
for different providers, which was the reason to have landmarks from
China, Argentina, etc. Can you please go back to my earlier emails on
this and put together a final comprehensive report with the main
takeaways, and also detailed data. This should describe ALL relevant
details of the telco networks and connectivity.

Rahul Nishant

unread,
Dec 15, 2013, 7:56:08 AM12/15/13
to ruralne...@googlegroups.com
Hi Sir,
Following are the results after doing the buffer size tests. The
buffersizes and the drainsizes are in KB. A detailed report will be sent
soon alongwith the traceroute results. These where the observations from
the buffersize tests
For downlink buffersizes (newdlbufData.png , newDownlinkBuffer.png):
-Airtel has comparatively higher buffersizes in rural areas as compared
to Idea and MTNL.
- The fact that the throughput of idea and MTNL are more than that of
Airtel (from thesis of Gautam and Gaurav), whereas the buffersizes of
these ISPs bein less than Airtel, shows unnecessary use of large
buffers in airtel networks.
-In urban area of Delhi Airtel has low buffersize than others. Also the
throughput of airtel from previous experiments shows high throughput for
Airtel. So large buffers can be seen to contribute to low throughput.
- The downlink buffers in 2G networks of Delhi are less than those in 3G
networks.

For uplink buffers (newulBuffData.png , newUplinkBuffer.png):
-The uplink buffers show similar sizes in the different locations.
- The buffers are smaller in 2G locations as compared to 3G locations of
Delhi and Jaipur, and Semi urban Dindori.

Following can be observed from the drainsizes (newDownlinkDrain.png,
newdldrainData.png, newUplinkDrain.png, newuldrainData.png):
- In 3G networks of Delhi, Dindori and Jaipur, it can be seen that the
downlink buffers drain in smaller chunks than the uplink buffers which
drain in larger chunks.
- The 2G locations have smaller drainsizes in both directions.
- Between ISPs, the drainsizes are mostly similar.

Currently I am working on the update mails, which needs changes to
accomodate the new tests and needs to be made more generic. Deadline :
20 Dec,2013

Thanks & regards,
Rahul Nishant



On 19.11.2013 07:33, Aaditeshwar Seth wrote:
newdlbufData.png
newDownlinkBuffer.png
newulbuffData.png
newUplinkBuffer.png
newdldrainData.png
newDownlinkDrain.png
newuldrainData.png
newUplinkDrain.png
Reply all
Reply to author
Forward
0 new messages