🐞 [Bug]: Grid proxy doesn't update used resources

[SalmaElsoly]

What happened?

when I try to list rentable nodes, there are some nodes which still has used resources but when I call statistics endpoint it shows the correct used_resources which is equal to system resources

which network/s did you face the problem on?

Main

Twin ID/s

No response

Version

No response

Node ID/s

1327, 1604, 1610

Farm ID/s

No response

Contract ID/s

No response

Relevant log output

https://gridproxy.grid.tf/nodes/1327
https://gridproxy.grid.tf/nodes/1327/statistics

[SalmaElsoly]

Investigations:

• checked how resources are added with contracts being created, when contract get created it's resources are added to the contract_resources there is triggers on that table that subtract or adds from used resources of the node based on the operation, i checked graphql running this

query MyQuery {
  nodeContracts(where: {nodeID_eq: 1327, state_eq: Deleted, resourcesUsed_isNull: true, deploymentData_not_contains: "network"}) {
    state
    nodeID
    contractID
    deploymentData
    resourcesUsed {
      cru
      hru
      mru
      sru
    }
  }
}

found that there are 13 vm contract that have null resources used for them, I thought that may be the cause that when updating the contract state there was no resource persisted to subtract but when I checked for normal updated nodes also found contracts that have no resources, I think there may be failure in the trigger still need to check db logs

[SalmaElsoly]

update:

after checking different stacks found that data are different so it's triggers problem since each stack has it's own db instance
https://gridproxy.grid.threefold.io/nodes/1327
https://gridproxy.grid.tf/nodes/1327
https://gridproxy.be.grid.tf/nodes/1327
https://gridproxy.sg.grid.tf/nodes/1327
https://gridproxy.us.grid.tf/nodes/1327