DynamoDB: Handling ProvisionedThroughputExceededException without increasing Read/Write Capacity - amazon-dynamodb

Question

Is there a good way to handle the ProvisionedThroughputExceededException without increasing the Read/Write capacity of the table? Ideally, I would like to decrease the throughput on the scripting side such that the exception won't be triggered.
function saveDataToDynamodb(data, scriptName){
let items = [];
var processItemsCallback = function (err, data) {
if (err) {
console.log(err);
setTimeout(() => {console.log('sleep for error')}, Math.floor(0.5 * 1000 + 1000));
}else {
if (Object.keys(data.UnprocessedItems).length > 0) {
//setTimeout(() => {console.log('sleep to wait')}, Math.floor(Math.random() * 1000 + 1000));
var params = {};
params.RequestItems = data.UnprocessedItems;
docClient.batchWrite(params, processItemsCallback);
}
}
};
data.forEach(station => {
let stationName = station.stationName
if (stationName == null) {
stationName = "null"
}
//console.log(station.stationId + "_" + stationName)
let pollutants = station.pollutants
let unique_set = new Set()
let aqiArr = [];
let sortedKey;
let timestamp;
pollutants.forEach(pollutant => {
timestamp = pollutant.utcMoment
sortedKey = station.stationId + "_" + timestamp + "_" + station.latitude + "_" + station.longitude + "_" + pollutant.pollutant;
if (!unique_set.has(sortedKey)) {
unique_set.add(sortedKey)
if (!scriptName.includes("Pollen")) { // calculate AQI if it's not pollen data
let polValue = processPollutant(pollutant.value, pollutant.pollutant);
let pollutantAQI;
if (pollutant.pollutant == "o3" && pollutant.avgInterval == "1h") {
pollutantAQI = aqiCalc.calcEpaAqi(polValue, pollutant.pollutant, true);
} else {
pollutantAQI = aqiCalc.calcEpaAqi(polValue, pollutant.pollutant, false);
}
//console.log("polValue: " + polValue + ", Pollutant value: " + pollutant.value + ", pollutant: " + pollutant.pollutant + ", AQI: " + pollutantAQI)
aqiArr.push(pollutantAQI);
}
items.push({
PutRequest: {
Item: {
"stationId": station.stationId,
"stationName": String(station.stationName),
"latitude": station.latitude,
"longitude": station.longitude,
"pollutant": pollutant.pollutant,
"avgInterval": pollutant.avgInterval,
"units": pollutant.units,
"initialUnits": pollutant.initialUnits,
"value": pollutant.value,
"sortedKey": sortedKey,
"timestamp": timestamp
}
}
});
if (items.length >= 24) {
let params = {
RequestItems: {
pollutantFullData: items
}
};
docClient.batchWrite(params, processItemsCallback);
items = []
}
}
});
if (aqiArr.length > 0) {
var finalAqi = getMaxAsInteger(aqiArr);
if (finalAqi > 500) {
finalAqi = 500;
}
items.push({
PutRequest: {
Item: {
"stationId": station.stationId,
"stationName": String(station.stationName),
"latitude": station.latitude,
"longitude": station.longitude,
"pollutant": "aqi",
"avgInterval": "1h",
"units": "ppb",
"initialUnits": "ppm",
"value": finalAqi,
"sortedKey": station.stationId + "_" + timestamp + "_" + station.latitude + "_" + station.longitude + "_aqi",
"timestamp": timestamp
}
}
});
}
if (items.length > 0) {
let params = {
RequestItems: {
pollutantFullData: items
}
};
docClient.batchWrite(params, processItemsCallback);
}
});
}
It seems as though the batchWrite operation is running asynchronously, such that the various threads are hitting the API endpoint at the same time, thus causing the error. How can the previous code be changed to resolve the error?