高并发-数据埋点，统计延时入库

背景：在活动运营时，通常我们会统计一下数据，比如 页面访问次数，按钮点击次数等等数据，当访问用户过多，高并发情况下，这些数据实时入库必然给数据库造成巨大的压力。所以，对高并发情况的下的数据统计，可采取延时入库的方案。

这里举个例子：

1.定义 实体类

/** *  * 日活统计表 *  *  */public class SaleDailyLogModel extends BaseDO {private static final long serialVersionUID = 264163675682797995L;public String getKey() {return this.actId + "_" + this.inviteCode + "_" + this.itemId;}public void addLog(SaleDailyLogModel log) {this.pageClickCount = this.pageClickCount + log.getPageClickCount();this.downloadCount = this.downloadCount + log.getDownloadCount();this.inviterPage = this.inviterPage + log.getInviterPage();this.inviteePage = this.inviteePage + log.getInviteePage();this.detailPage = this.detailPage + log.getDetailPage();this.shareClick = this.shareClick + log.getShareClick();this.sharePage = this.sharePage + log.getSharePage();this.download = this.download + log.getDownload();this.inviterBuy = this.inviterBuy + log.getInviterBuy();this.inviteeBuy = this.inviteeBuy + log.getInviteeBuy();}public void clearLog() {this.pageClickCount = 0;this.downloadCount = 0;this.inviterPage = 0;this.inviteePage = 0;this.detailPage = 0;this.shareClick = 0;this.sharePage = 0;this.download = 0;this.inviterBuy = 0;this.inviteeBuy = 0;this.timestamp = System.currentTimeMillis();}private Long id;private Long actId; // 活动idprivate String inviteCode = "";private Long itemId;private Integer pageClickCount = 0; // 页面点击数量private Integer downloadCount = 0; // 下载数量private Date date; // 日期private long timestamp = System.currentTimeMillis();private int inviterPage;private int inviteePage;private int detailPage;private int shareClick;private int sharePage;private int download;private int inviterBuy;private int inviteeBuy;public Long getId() {return id;}public void setId(Long id) {this.id = id;}public Long getActId() {return actId;}public void setActId(Long actId) {this.actId = actId;}public Integer getPageClickCount() {return pageClickCount;}public void setPageClickCount(Integer pageClickCount) {this.pageClickCount = pageClickCount;}public Integer getDownloadCount() {return downloadCount;}public void setDownloadCount(Integer downloadCount) {this.downloadCount = downloadCount;}public Date getDate() {return date;}public void setDate(Date date) {this.date = date;}public String getInviteCode() {return inviteCode;}public void setInviteCode(String inviteCode) {this.inviteCode = inviteCode;}public Long getItemId() {return itemId;}public void setItemId(Long itemId) {this.itemId = itemId;}public int getInviterPage() {return inviterPage;}public void setInviterPage(int inviterPage) {this.inviterPage = inviterPage;}public int getInviteePage() {return inviteePage;}public void setInviteePage(int inviteePage) {this.inviteePage = inviteePage;}public int getDetailPage() {return detailPage;}public void setDetailPage(int detailPage) {this.detailPage = detailPage;}public int getShareClick() {return shareClick;}public void setShareClick(int shareClick) {this.shareClick = shareClick;}public int getSharePage() {return sharePage;}public void setSharePage(int sharePage) {this.sharePage = sharePage;}public int getDownload() {return download;}public void setDownload(int download) {this.download = download;}public int getInviterBuy() {return inviterBuy;}public void setInviterBuy(int inviterBuy) {this.inviterBuy = inviterBuy;}public int getInviteeBuy() {return inviteeBuy;}public void setInviteeBuy(int inviteeBuy) {this.inviteeBuy = inviteeBuy;}public long getTimestamp() {return timestamp;}public void setTimestamp(long timestamp) {this.timestamp = timestamp;}}

2.实现类：

 @Autowired    protected SaleLogManager saleLogManager;        private Map<String, SaleDailyLogModel> dailyLogMap = new ConcurrentHashMap<String, SaleDailyLogModel>();        @Overridepublic void saveSaleDailyLog(SaleDailyLogModel saleDailyLogModel) {    if(dailyLogMap.get(saleDailyLogModel.getKey()) == null) {    dailyLogMap.put(saleDailyLogModel.getKey(),saleDailyLogModel);    }    SaleDailyLogModel dailyLog = dailyLogMap.get(saleDailyLogModel.getKey());    synchronized (dailyLog) {    dailyLog.addLog(saleDailyLogModel);//if也可设置一个固定的时间段入库if(System.currentTimeMillis() - dailyLog.getTimestamp() > ((int) (Math.random() * 60000 + 30000))) {try {saleLogManager.saveSaleDailyLog(dailyLog);//入库dailyLog.clearLog();} catch (SQLException e) {log.error("save daily log error:" , e);}}}    }