Task #2726
openLoad repos with CMIP5 data
0%
Updated by Chengyu Fan over 9 years ago
Here is the design:
- 1. produce the corresponding TLV file for each data file (should create a tool based on ndn-cxx, this tool reads the NDN name and generates the NDN TLV file)
- 2. utilize repo-ng TCP bulk insertion protocol to read the TLV file (e.g., nc localhost 7376 < x.tlv)
Noe that the TLV file generation is slow when we need to sign each content segment (about 4 minutes for a 1.2GB file when we use 8K data segment), we may use other efficient method in the future.
If you have better designs, please let me know.
@susmit shannigrahi, if there is no objections, please tell me how to get the NDN name for each data file? I think the best way is to read the translator result, because we just need to add tools if we need to use another method to load the data.
Updated by susmit shannigrahi over 9 years ago
Could you put the code in gerrit please? I need to take a look to see how to integrate this. If you want, you can assign this to me.
Updated by Chengyu Fan over 9 years ago
I only have the TLV file generation code? I'll upload it soon
Updated by Chengyu Fan over 9 years ago
Please refer http://gerrit.named-data.net/#/c/1974/1 for the TLV file generation code