1.概述
前面在《ElasticSearch實戰－入門》中給大家分享如何搭建這樣一個叢集，在完成叢集的搭建後，今天給大家分享如何實現對應的業務功能模組，下面是今天的分享內容，目錄如下所示：
- 編碼實踐
- 效果預覽
- 總結
2.編碼實踐
由於 ES 叢集支援 Restful 介面，我們可以直接通過 Java 來呼叫 Restful 介面來查詢我們需要的資料結果，並將查詢到的結果在在我們的業務介面視覺化出來。我們知道在 ES 叢集的 Web 管理介面有這樣一個入口，如下圖所示：
我們可以在此介面的入口中拼接 JSON 字串來查詢我們想要的結果，下面，我們通過 Java 的 API 去呼叫 Restful 介面來查詢我們想要的結果。
2.1 字串拼接實現
接著，我們去實現要查詢的核心程式碼，具體內容實現如下所示：
public String buildQueryString(Map<String, Object> param) throws ParseException { SimpleDateFormat dfs = new SimpleDateFormat("yyyy-MM-dd HH:mm:ss"); StringBuilder builder = new StringBuilder("{\"query\":{\"bool\":{\"must\":["); if (param.get("msgType") != null) { Integer msgType = (int) param.get("msgType") == 0 ? 2 : 1; builder.append("{\"term\":{\"msg_type\":").append(msgType).append("}}"); }if (param.get("start") != null && param.get("end") != null) { String start = String.valueOf(dfs.parse(param.get("start").toString()).getTime()).substring(0, 10); String end = String.valueOf(dfs.parse(param.get("end").toString()).getTime()).substring(0, 10); builder.append(",{\"range\":{\"itime\":{\"from\":" + start + ",\"to\":" + end + "}}}"); } if (param.get("receiverValue") != null) { builder.append(",{\"wildcard\":{\"receiver_value\":\"*").append(param.get("receiverValue")).append("*\"}}"); } builder.append("],\"must_not\":[],\"should\":[]}}"); builder.append(",\"sort\":[{\"itime\":\"desc\"}],\"facets\":{}"); builder.append(",\"from\": ").append(param.get("startIndex")).append(",\"size\": ").append(param.get("pageSize")).append("}"); LOG.info("API Query -> " + builder.toString()); return builder.toString(); }
2.2 查詢實現核心程式碼
接著是實現查詢的核心程式碼，具體內容實現如下所示：
public SerachResponse<ApiSent> querySent(Map<String, Object> param) { SerachResponse<ApiSent> search_result = null; try { long time = System.currentTimeMillis(); ResponseWrapper wrapper = httpUtils.sendJson(configService.loadConfig(Configure.API_SENT), buildQueryString(param)); if (wrapper.responseCode == HttpStatus.SC_OK) { search_result = _gson.fromJson(wrapper.responseContent, new TypeToken<SerachResponse<ApiSent>>() { }.getType()); LOG.info(String.format("API query ES spent time=%sms", (System.currentTimeMillis() - time))); return search_result; } else { LOG.info(String.format("API query ES spent time=%sms", (System.currentTimeMillis() - time))); LOG.error(String.format("api sent request es server response not 200,response=%s,exception=%s", wrapper.error, wrapper.exceptionString)); } } catch (Exception ex) { LOG.error(String.format("parsed es sent data exception.", ex)); } return search_result; }
- Configure類
public class Configure { public static final String API_SENT = "API_SENT"; ｝
2.3 DAO層獲取 ES 叢集的連線資訊
public class ConfigService { private static Log logger = LogFactory.getLog(ConfigService.class); @Autowired private ConfigDao configDao; @Cacheable("sysConfigCache") public String loadConfig(String type) { String value = configDao.getConfig(type); logger.info(String.format("Load Config,type=%s,value=%s", type, value)); return value; } }
- ConfigDao介面
public interface ConfigDao { String getConfig(String type); }
其對應的實現內容如下所示：
<select id="getConfig" parameterType="String" resultType="String"> select value from t_system_config where type=#{type} </select>
DB庫儲存的 ES 連線資訊，如下圖所示：
2.4 HTTP 介面的實現程式碼
關於 HttpUtils 的程式碼實現較為簡單，這裡直接附上程式碼的實現內容，如下所示：
- IHttpUtils 介面
public interface IHttpUtils { public ResponseWrapper sendJson(String url, String content); }
- HttpUtils 類實現介面
public class HttpUtils implements IHttpUtils { private static Logger LOG = Logger.getLogger(HttpUtils.class.getName()); protected static Gson _gson = new Gson(); protected void initSSL() { try { TrustManager[] tmCerts = new javax.net.ssl.TrustManager[1]; tmCerts[0] = new SimpleTrustManager(); javax.net.ssl.SSLContext sc = javax.net.ssl.SSLContext.getInstance("SSL"); sc.init(null, tmCerts, null); javax.net.ssl.HttpsURLConnection.setDefaultSSLSocketFactory(sc.getSocketFactory()); HostnameVerifier hv = new SimpleHostnameVerifier(); HttpsURLConnection.setDefaultHostnameVerifier(hv); } catch (Exception e) { LOG.error("init SSL exception.", e); } } @Override public ResponseWrapper sendJson(String url, String content) { return sendJson(url, content, null); } @Override public ResponseWrapper sendJson(String url, String content, String authCode) { return sendRequest(url, content, METHOD_POST, authCode, CONTENT_TYPE_JSON); } public ResponseWrapper sendRequest(String url, String content, String method, String authCode, String contentType) { LOG.info("Send request to - " + url + ", with content - " + content); HttpURLConnection conn = null; OutputStream out = null; StringBuffer sb = new StringBuffer(); cn.jpush.utils.ResponseWrapper wrapper = new ResponseWrapper(); try { if (StringUtils.isSSL(url)) { initSSL(); } if (METHOD_GET.equals(method)) { if (!Strings.isNullOrEmpty(content)) url += "?" + content; } URL aUrl = new URL(url); wrapper.address = aUrl.getHost(); conn = (HttpURLConnection) aUrl.openConnection(); conn.setConnectTimeout(DEFAULT_CONNECTION_TIMEOUT); conn.setReadTimeout(DEFAULT_SOCKET_TIMEOUT); conn.setUseCaches(false); conn.setRequestMethod(method); conn.setRequestProperty("Connection", "Keep-Alive"); conn.setRequestProperty("Accept-Charset", CHARSET); conn.setRequestProperty("Charset", CHARSET); conn.setRequestProperty("Authorization", authCode); conn.setRequestProperty("Send-Source", "portal"); conn.setRequestProperty("Content-Type", contentType); if (METHOD_POST.equals(method)) { conn.setDoOutput(true); byte[] data = content.getBytes(CHARSET); conn.setRequestProperty("Content-Length", String.valueOf(data.length)); out = conn.getOutputStream(); out.write(data); out.flush(); } else { conn.setDoOutput(false); } int status = conn.getResponseCode(); InputStream in = null; if (status == 200) { in = conn.getInputStream(); } else { in = conn.getErrorStream(); } InputStreamReader reader = new InputStreamReader(in, CHARSET); char[] buff = new char[1024]; int len; while ((len = reader.read(buff)) > 0) { sb.append(buff, 0, len); } String responseContent = sb.toString(); wrapper.responseCode = status; wrapper.responseContent = responseContent; String quota = conn.getHeaderField(RATE_LIMIT_QUOTA); String remaining = conn.getHeaderField(RATE_LIMIT_Remaining); String reset = conn.getHeaderField(RATE_LIMIT_Reset); wrapper.setRateLimit(quota, remaining, reset); if (status == 200) { LOG.debug("Succeed to get response - 200 OK"); LOG.debug("Response Content - " + responseContent); } else if (status > 200 && status < 400) { LOG.warn("Normal response but unexpected - responseCode:" + status + ", responseContent:" + responseContent); } else { LOG.warn("Got error response - responseCode:" + status + ", responseContent:" + responseContent); switch (status) { case 400: LOG.error("Your request params is invalid. Please check them according to error message."); wrapper.setErrorObject(); break; case 401: LOG.error("Authentication failed! Please check authentication params according to docs."); wrapper.setErrorObject(); break; case 403: LOG.error("Request is forbidden! Maybe your is listed in blacklist?"); wrapper.setErrorObject(); break; case 410: LOG.error("Request resource is no longer in service. Please according to notice on official website."); wrapper.setErrorObject(); case 429: LOG.error("Too many requests! Please review your request quota."); wrapper.setErrorObject(); break; case 500: case 502: case 503: case 504: LOG.error("Seems encountered server error. Maybe is in maintenance? Please retry later."); break; default: LOG.error("Unexpected response."); } } } catch (SocketTimeoutException e) { if (e.getMessage().contains(KEYWORDS_READ_TIMED_OUT)) { LOG.error(KEYWORDS_READ_TIMED_OUT, e); } wrapper.exceptionString = e.getMessage(); } catch (IOException e) { LOG.error(KEYWORDS_CONNECT_TIMED_OUT, e); wrapper.exceptionString = e.getMessage(); } finally { if (null != out) { try { out.close(); } catch (IOException e) { LOG.error("Failed to close stream.", e); } } if (null != conn) { conn.disconnect(); } } LOG.info(String.format("Send Response to - %s, Response Wrapper - %s", url, wrapper)); return wrapper; } }
3.截圖預覽
下面給大家附上一張業務介面視覺化的資料結果預覽圖，如下圖所示：
上圖為我傳送的測試資料，通過收集模組將我傳送的資料收集並儲存到 ES 叢集，通過介面程式碼將這部分資料視覺化到業務介面進行展示。
4.總結
總體來說，ES 叢集從搭建部署到編碼實現都較為簡單，在使用 JSON 字串拼接查詢時需要細心點，後續有時間可以為大家分享下 ES 的查詢的效率，及其他方面的效能指標。
5.結束語
這篇部落格就和大家分享到這裡，如果大家在研究學習的過程當中有什麼問題，可以加群進行討論或傳送郵件給我，我會盡我所能為您解答，與君共勉！