Skip to main content

Introduction

The purpose of Tracker is to download the data from rollup to ensure the consistency of data sets’ order.

Implementation

func (t *Tracker) jobTxsPullFromChain() {
log.Info("job txs pull running...")
defer func() {
log.Info("job txs pull done")
time.Sleep(5 * time.Second)
t.isSynced = true
}()

// get all unprocessed txs
lastTxID, err := t.store.GetConstant(schema.LastProcessArTxIdKey)

ids := reverseIDs(MustFetchTxIdsByNativeMethod(t.qryTags, lastTxID, t.arClient))

// process txs
arTxsCounts := 0
bizTxsCounts := 0

for _, id := range ids {
owner, timestamp, height, data := t.mustFetchDataByID(id)

// 1. process
// Only execute transactions where the owner and arOwner are the same or where the arOwner is ""
num := 0
if t.arOwner == "" || t.arOwner == owner {
num = t.processTxs(id, owner, timestamp, height, data)
}

// 2. update last process ar tx id to store
if err := t.store.UpdateLastProcessArTxId(id); err != nil {
log.Error("update last process ar tx id store error", "err", err)
panic(err)
}

log.Info("process txs", "arTxId", id, "number", num) // If num = 0 means the arTx is not part of the transaction on the arOwner chain
bizTxsCounts += num
arTxsCounts++
}

}

Note:

  • tracker pull AR transactions uploaded by rollup from Arweave in the form of a scheduled task.
  • After obtaining AR transactions, tracker extracts the data of the transactions one by one in the submission order of the transaction, and parses the data into the form of the pooltx in rollup, which contains the data items externally submitted to rollup.
  • After obtaining the dataItem, tracker will encapsulate the dataItem as a SubscribeTransaction and then submit it to the channel. Other modules that subscribe to the channel can sequentially access the data drawn from the chain and uploaded by rollup.