2rd Community Review of Filscanteam Allocator

[FilscanOfficial]

Allocator application filecoin-project/notary-governance#1033

Our second batch of 5PiB is about to be consumed.
Two new clients were added in this round:
FilscanOfficial/filscan-backend#23
FilscanOfficial/filscan-backend#21
There is also an old client that continues to supply:
FilscanOfficial/filscan-backend#17

Compared with the first batch, we have made the following improvements and report them here:

1. Disclose all client application information on GitHub to provide evidence for due diligence;
2. When the client has changed information, let them change the application in time;
3. In order to improve the retrieval rate, it is recommended that the client stop using DDO to issue orders, and require all to improve the retrieval rate;

Since the demand of the three clients that are still in supply has not been met, and we also hope to attract more clients to join filecoin, so this round hopes to supply 10PiB, thank you very much!

[filecoin-watchdog]

@FilscanOfficial This is the 2nd review, not the 3rd :)
Allocator Application
Compliance Report
1st Review
1st Review score: 5PiB

5 PiB granted to clients:

Client Name	DC	status
ABGAasia	1.13PiB	New
DuitangInformation Technology (Shanghai) Co., Ltd.	3.67PiB	New
FileDriveLabs	0.2PiB	Existing

ABGAasia

• KYB completed via email.
• Dataset already stored on Filecoin: Issue #872.
• Insufficient information on data preparation (missing from the form; allocator did not request it).
• SPs used for deals align with the provided or updated list.
• Majority of SPs located in the same region.
• Low retrieval rate.

Duitang Information Technology (Shanghai) Co., Ltd

• KYB completed via email.
• 200 TiB granted for testing after KYB/KYC.
• SPs used for deals align with the provided or updated list.
• Good retrieval and replication performance.

FileDriveLabs

• SPs used for deals align with the provided or updated list.
• Two SPs sealed excessive data (f02984331, f02883857).

[FilscanOfficial]

@filecoin-watchdog
I am very happy to receive your reply. Thank you for pointing out my error. I have corrected the third time to the second time.
For my current shortcomings, I will make dissolution and improvement measures here:
FilscanOfficial/filscan-backend#21

• The dataset has been stored on Filecoin: [DataCap Application] Asia Blockchain Gaming Alliance filecoin-plus-large-datasets#872.
  I asked the client and he explained that since it has been a long time, there have been personnel changes and new data has been generated, so I applied again. Before the next round of approval, I will ask them to provide more data samples for review.
  

• Insufficient information on data preparation (missing from the form; allocator did not request it).
  This is my mistake. I ignored the strict review of private data sets. I will improve it in the future through the following means

1. Download data from the IP specified by the customer for review.
2. Ask the customer to provide more data samples when applying for connection.
3. Ask the customer to provide more data samples for review by email.

• Majority of SPs located in the same region.
  When I applied for allcator, I did not restrict the customer's region, because I think that assigning different SPs can ensure data security. Of course, we can also suggest customers to try to assign regions in the future.

• Low retrieval rate.
  For SPs with poor retrieval results, I will ask the client to explain and elaborate on the improvement plan (in the first round of testing, one SP was not retrieved, and I followed up the reason in time)
  
  
  Client solved their problems and retrieval improved，I will continue to pay attention to the retrieval rate in the future.
  

FilscanOfficial/filscan-backend#17

• Two SPs sealed excessive data (f02984331, f02883857).
  In this round, I first provided 200T for testing, and they packed 40T for each of the 5 SPs, which met the requirements. Since he is an old customer, the report contains historical data, so it shows that the 2 SPs packed too much data. I controlled this round more strictly, so after testing the average distribution of the 5 SPs, I still asked the other party to explain the subsequent distribution plan, and I could only approve it after the other party responded. (The SP that does not show the search is the DDO node of the previous round)
  
  

Finally, thank you again for your hard work in helping me review compliance. In the future, I will become more and more standardized and strive to introduce more outstanding customers to join filecoin. I wish you a happy work!