search

CirclesUBI / circles-subgraph

Circles subgraph for The Graph
https://thegraph.com/explorer/subgraph/circlesubi/circles-ubi
GNU Affero General Public License v3.0
5 stars 5 forks source link

Subgraph fails to sync #52

Closed llunaCreixent closed 2 years ago

llunaCreixent commented 3 years ago

The circlesubi subgraph stopped syncing on block 19,001,246. No errors in the logs. Now when I try to deploy starting from block 12,529,458 it fails with these warnings: 

Unexpected RPC error, error: Transport("Unexpected response status code: 413 Payload Too Large"), component: BlockStream
Trying again after eth_getLogs RPC call failed (attempt #10) with result Err(Transport("Unexpected response status code: 413 Payload Too Large")), component: BlockStream.

The deployment works if I deploy from a few blocks before it failed, but the data from previous blocks is lost.

llunaCreixent commented 3 years ago

Some Questions:

leoyvens commented 3 years ago

The hosted service closely follows graph node master. We have had previous reports of "payload to large" https://github.com/graphprotocol/graph-node/issues/2616. I noticed that your subgraph is not updated to the latest graph-cli/-ts and to apiVersion 0.0.5, that might be unrelated to these issues but it is strongly recommended.

llunaCreixent commented 3 years ago

I updated graph-cli/-ts (see https://github.com/CirclesUBI/circles-subgraph/pull/53) to the latest and the subgraph it's not successfully synced in the Hosted Service (see: https://thegraph.com/hosted-service/subgraph/circlesubi/circles?selected=logs).

leoyvens commented 3 years ago

@llunaCreixent are you referring to the current or pending version?

llunaCreixent commented 3 years ago

I was referring to the pending version, it stopped after having 199 entities. Then, I deployed from a later block, and started syncing again. Then, I deployed from the first block where Circles txs exist. Although now it's syncing, we are not sure if it indexed the first txs, and from time to time we can see these warnings in the logs:

Trying again after eth_getBlockByNumber RPC call failed (attempt #15) with result Err(Transport error: Unexpected response status code: 502 Bad Gateway), component: BlockStream
leoyvens commented 3 years ago

@llunaCreixent We have rolled out batching of eth_getLogs, so the 413 errors should be fixed.

llunaCreixent commented 2 years ago

We updated the subgraph version and now this issue in not relevant anymore. Thanks for your help @leoyvens !!