Request for Information about Disputes #GDR-48

Governance & GIPs
1

The Arbitrators are contacting Indexer address 0xdecba5154aab37ae5e381a19f804f3af4d1bcbb5 and fisherman 0x0a015d9e17d3978164486473a2ddbb654418db07 for a new dispute filed in the protocol.

Dispute ID: 0x93d7057d16960f0ce69410888cfa19318f90412f1e7d7e37336c9e0798abde33
Allocation: 0x4698ec86efcdf91c4bf341effda61c9c5c43e7dc
Subgraph: QmVcBWJZQNYKJpd1Vx9BLrgjvxnmuev7JhD2HvLVn4XBy6

Fisherman, could you share the insights or data you gathered that led to you filing the dispute? Please provide all relevant information and records about the open dispute. This will likely include POIs generated for the affected subgraph(s).

About the Procedure

The Arbitration Charter regulates the arbitration process. You can find it in Radicle project ID rad:git:hnrkrhnth6afcc6mnmtokbp4h9575fgrhzbay or at GIP-0009: Arbitration Charter.

Please use this forum for all communications.

Arbitration Team

2

Dispute Summary

Field Value
Dispute ID 0x93d7057d16960f0ce69410888cfa19318f90412f1e7d7e37336c9e0798abde33
Allocation ID 0x4698ec86efcdf91c4bf341effda61c9c5c43e7dc
Subgraph QmVcBWJZQNYKJpd1Vx9BLrgjvxnmuev7JhD2HvLVn4XBy6
Network avalanche
POI Block Number 73,333,339
Closed At Epoch 1103
Charter Type Legacy (GIP-0009)

Disputed POI

Field Value
Indexer 0xdecba5154aab37ae5e381a19f804f3af4d1bcbb5
Operator 0xc4e706f488751cecc0a184e5a7acff5f35b0d5bb
Submitted POI 0xe46fb807bdc17d6470885596d87544a6753ec3c07aa15d019a523cded9a0f9cf
Expected POI 0xce1fd638ad9aa4f7b9f7935313ce6df78582b40470e50318c957d908ef3037f2
Result Mismatch

Corroborating Evidence

Two other indexers closed allocations on the same subgraph at the same POI block (73,333,339) and both submitted POIs that match my graph-node’s calculation:

Indexer Submitted POI My Calculated POI Result
0xfc842f81490dcb37e82d416b2d28327dfb24ba9a 0xc9735e98a1f885846afda98089256744c1809894d5634e14a08c59f67f08b459 0xc9735e98a1f885846afda98089256744c1809894d5634e14a08c59f67f08b459 Match
0xf00f7157fa8fd0420b87956d46058a16b2f23adc 0x3610f2d31f49ec2c505e6ae4b2999957a173c7e02bece8ed10e80e5e9564358e 0x3610f2d31f49ec2c505e6ae4b2999957a173c7e02bece8ed10e80e5e9564358e Match

Both indexers independently computed the same POI as my node at this block. The disputed indexer is the only one whose POI diverges.

Cross-Subgraph Analysis

The disputed indexer also indexes two other Avalanche subgraphs: QmeuzzdgrN9mjAYQqVENLDRWtREV7wDMGVEthhg1sY4xdu and QmWFbXcvxqD39BjzJ72WwF4fe8ZGH2Z25ZcrF7onCPPBpr. I verified their POIs on these subgraphs and found a correlated pattern of mismatches.

The disputed indexer mismatched on multiple subgraphs throughout a block range, then recovered on all of them simultaneously:

POI Block QmVcBWJZQNYKJpd1Vx9BLrgjvxnmuev7JhD2HvLVn4XBy6 (this dispute) QmeuzzdgrN9mjAYQqVENLDRWtREV7wDMGVEthhg1sY4xdu QmWFbXcvxqD39BjzJ72WwF4fe8ZGH2Z25ZcrF7onCPPBpr
72,955,253 match – match
73,333,339 Mismatch – Mismatch
74,472,074 Mismatch Mismatch Mismatch
75,381,882 Match Match Match

On QmeuzzdgrN9mjAYQqVENLDRWtREV7wDMGVEthhg1sY4xdu, the disputed indexer is the only indexer with any mismatch. Three other indexers (0x3e1536fc83cd5bed83a521a26034ff3e59c6a7c4, 0x8cc22436ba6f07a4d5dd2043e3109267eee5aab8, 0x0a015d9e17d3978164486473a2ddbb654418db07 ← self) all matched across 7 allocations spanning blocks 72,836,707 to 75,811,870.

On QmWFbXcvxqD39BjzJ72WwF4fe8ZGH2Z25ZcrF7onCPPBpr, four other indexers (0x53dbbc9d916b1840a2c4c26b150ba2e13f36e10f, 0x0a015d9e17d3978164486473a2ddbb654418db07, 0xf92f430dd8567b0d466358c79594ab58d919a6d4, 0x3e1536fc83cd5bed83a521a26034ff3e59c6a7c4) all matched at blocks surrounding the disputed indexer’s mismatches. 0x53dbbc9d916b1840a2c4c26b150ba2e13f36e10f matched at block 74,537,393 – immediately after the mismatch at 74,472,074.

I do not believe that 3 seemingly independent subgraphs with different code, different mappings, different entities, etc… simply break and then recover in unison, if not for infra issues, unless there is another factor I have not considered? The shared variable across all three subgraphs with mismatching PoI’s is the indexer’s own infrastructure/RPC. This pattern indicates to me that the mismatches originate from the indexer’s setup rather than from any individual subgraph.

GDR-49, covers a separate Horizon-era allocation by the same indexer on the same subgraph at block 74,472,074.

Full Verification Data

All closed allocations on QmVcBWJZQNYKJpd1Vx9BLrgjvxnmuev7JhD2HvLVn4XBy6 verified against my graph-node:

Indexer Block Submitted POI My Calculated POI Result
0xf00f7157fa8fd0420b87956d46058a16b2f23adc 77,114,939 0xdc2fb880...988993 0xdc2fb880...988993 Match
0x2f09092aacd80196fc984908c5a9a7ab3ee4f1ce 77,114,939 0x78b90da2...3fb7ad 0x78b90da2...3fb7ad Match
0xf92f430dd8567b0d466358c79594ab58d919a6d4 77,039,126 0xa303d679...3a547c 0xa303d679...3a547c Match
0x35917c0eb91d2e21bef40940d028940484230c06 76,732,196 0x5d771d61...5bf8a7 0x5d771d61...5bf8a7 Match
0xfc842f81490dcb37e82d416b2d28327dfb24ba9a 76,238,053 0xbebfa124...8bc0d8 0xbebfa124...8bc0d8 Match
0x35917c0eb91d2e21bef40940d028940484230c06 76,238,053 0x35822fe7...e7a276 0x35822fe7...e7a276 Match
0xdecba5154aab37ae5e381a19f804f3af4d1bcbb5 76,167,145 0xb90bffcf...03c775 0xb90bffcf...03c775 Match
0xfc842f81490dcb37e82d416b2d28327dfb24ba9a 76,095,560 0x740ad7e5...4e3730 0x740ad7e5...4e3730 Match
0x35917c0eb91d2e21bef40940d028940484230c06 76,095,560 0xf708d520...e34264 0xf708d520...e34264 Match
0xfc842f81490dcb37e82d416b2d28327dfb24ba9a 75,594,156 0x0014bde9...712827 0x0014bde9...712827 Match
0xdecba5154aab37ae5e381a19f804f3af4d1bcbb5 75,381,882 0x1b1b097f...79e13f 0x1b1b097f...79e13f Match
0x35917c0eb91d2e21bef40940d028940484230c06 75,239,491 0xb91dbfbc...f216e9 0xb91dbfbc...f216e9 Match
0xf92f430dd8567b0d466358c79594ab58d919a6d4 75,239,491 0xaff30f4b...1aaa13 0xa3d1867d...607b4e Mismatch
0xf00f7157fa8fd0420b87956d46058a16b2f23adc 75,174,859 0xf0fea642...71868a 0xf0fea642...71868a Match
0x2f09092aacd80196fc984908c5a9a7ab3ee4f1ce 75,174,859 0x7f2e3495...a85e49 0x7f2e3495...a85e49 Match
0x35917c0eb91d2e21bef40940d028940484230c06 75,174,859 0xcdb7683f...cfe67f 0xcdb7683f...cfe67f Match
0xdecba5154aab37ae5e381a19f804f3af4d1bcbb5 74,472,074 0xc3304988...fcaf53 0x0e92763b...c4c3ef Mismatch
0xdecba5154aab37ae5e381a19f804f3af4d1bcbb5 74,282,550 0xbc5bcba3...600872 0xbc5bcba3...600872 Match
0xdecba5154aab37ae5e381a19f804f3af4d1bcbb5 74,154,903 0xa2bcfef4...8adcaa 0xa2bcfef4...8adcaa Match
0xfc842f81490dcb37e82d416b2d28327dfb24ba9a 73,780,735 0x6bb41cc7...9bc773 0x6bb41cc7...9bc773 Match
0xf00f7157fa8fd0420b87956d46058a16b2f23adc 73,461,249 0xe8722e2a...04b2b9 0xe8722e2a...04b2b9 Match
0x2f09092aacd80196fc984908c5a9a7ab3ee4f1ce 73,461,249 0x658d1135...060a30 0x658d1135...060a30 Match
0xf92f430dd8567b0d466358c79594ab58d919a6d4 73,397,298 0x0dfb65e8...f08b2d 0x0dfb65e8...f08b2d Match
0xfc842f81490dcb37e82d416b2d28327dfb24ba9a 73,333,339 0xc9735e98...08b459 0xc9735e98...08b459 Match
0xf00f7157fa8fd0420b87956d46058a16b2f23adc 73,333,339 0x3610f2d3...64358e 0x3610f2d3...64358e Match
0xdecba5154aab37ae5e381a19f804f3af4d1bcbb5 73,333,339 0xe46fb807...a0f9cf 0xce1fd638...3037f2 Mismatch
0xdecba5154aab37ae5e381a19f804f3af4d1bcbb5 72,955,253 0x5f20a938...76a08d 0x5f20a938...76a08d Match
0xfc842f81490dcb37e82d416b2d28327dfb24ba9a 72,895,891 0xcf41bfa5...70f341 0xcf41bfa5...70f341 Match
0x2f09092aacd80196fc984908c5a9a7ab3ee4f1ce 72,895,891 0xc71c7484...997218 0xc71c7484...997218 Match
0x35917c0eb91d2e21bef40940d028940484230c06 72,601,656 0xab79be77...e8666c 0xab79be77...e8666c Match
0xdecba5154aab37ae5e381a19f804f3af4d1bcbb5 72,601,656 0x43db2a4a...97c1ec 0x43db2a4a...97c1ec Match
0x3f74870f80ff7449fe4c6ff257da5fa72734c970 72,089,542 0x3bb8031e...e5adab 0x3bb8031e...e5adab Match
0xdec965f0604125be05cd8a136c85d02ef344d61a 72,089,542 0x77bb5742...77f52e 0x77bb5742...77f52e Match
0xdeb712db301285ed483ef9e02dd08a1980f273f1 72,089,542 0x08266ecb...4eb24d 0x08266ecb...4eb24d Match
0x6f3ce93a09f30f18d728d2364268b5fe9444b89e 72,089,542 0x84a199de...f80317 0x84a199de...f80317 Match
0xf92f430dd8567b0d466358c79594ab58d919a6d4 72,032,972 0xb16ed24a...fb65de 0xb16ed24a...fb65de Match
0xf00f7157fa8fd0420b87956d46058a16b2f23adc 71,709,769 0x70c82855...cbb143 0x70c82855...cbb143 Match
0xdecba5154aab37ae5e381a19f804f3af4d1bcbb5 71,709,769 0x3032e418...76bd24 0x3032e418...76bd24 Match

Closing Transaction

Field Value
Arbitrum Block Hash 0x1b2788c940f9c5aa2882ff5b7a1c700c270440d4d4b503141b63688da78add45
Arbiscan Link View on Arbiscan

Final note

I am happy to provide any additional information the Arbitration Team requires, including the full POI verification data for the supporting subgraphs referenced in the cross-subgraph analysis.

3

Thanks @badcoffee for another stellar investigation and report. Really appreciate the effort you put in.

After reviewing the submitted evidence and the additional analysis performed, the fisherman’s conclusions appear correct. The observed POI mismatches are most plausibly explained by a temporary data-ingest issue on the disputed indexer’s side. Unfortunately we haven’t heard from the indexer and their endpoint seems to be down so there is also no way of verifying their current POIs. However as the fisherman points out the indexer has posted valid POIs after the mismatched ones suggesting that whatever issue they were experiencing was already solved.

Following guidance from section 13 of the Arbitration Charter (graph-improvement-proposals/gips/0085-horizon-arbitration-charter.md at main · graphprotocol/graph-improvement-proposals · GitHub), the suggested outcome is to resolve the dispute as a draw and give the indexer 7 days to address the issue. It’s worth noting that both GDR-48 and GDR-49 correspond to the same indexer showcasing the same behavior but in different periods of time, three weeks or so apart. Given the indexer has fixed the issue of their own merit the arbitrators understand the indexer did not act in bad faith.

We’ll be posting the transaction shortly. Once again @badcoffee thank you for your continued work.

4

Transaction posted here: Arbitrum One Transaction Hash: 0x4c132b20b2... | Arbitrum One