Solr在5.0版本后,不再提供war包部署的方式,取而代之的是内置了jetty服务,
但是我们发现其中并没有内置任何安全性相关检查,任何人如果知道了我们的外网地址就能直接访问并修改其中的索引。经过查找可以使用jetty的方式来限制web访问。
solr/server/solr-webapp/WEB-INF/web.xml中增加以下字段:
<security-constraint> <web-resource-collection> <web-resource-name>solr</web-resource-name> <url-pattern>/</url-pattern> </web-resource-collection> <auth-constraint> <role-name>solr_admin</role-name> <role-name>admin</role-name> </auth-constraint> <login-config> <auth-method>BASIC</auth-method> <realm-name>Solr Admin</realm-name> </login-config> </security-constraint>
配置验证方式BASIC(用户名密码的方式)。
solr/server/etc/jetty.xml中增加Call标签:
<Call name="addBean"> <Arg> <New class="org.eclipse.jetty.security.HashLoginService"> <Set name="name">Solr Admin</Set> <Set name="config"> /Users/mazhiqiang/develop/tools/solr-5.5.0/server/etc/realm.properties </Set> <Set name="refreshInterval">0</Set> </New> </Arg> </Call>
config中指定密码文件的路径,可以在其中使用<SystemProperty>来共同组合路径,例如配置了环境变量的情况下,可以使用下面的方式:
<Set name="config"><SystemProperty name="jetty.home" default="."/>/etc/realm.properties</Set>
而指定的realm.properties就是相关密码文件了:
admin:xxxx,solr_admin
设置完成,重新启动solr即可,如果不输入用户名和密码,无法登陆成功:
![](http://dl2.iteye.com/upload/attachment/0119/4232/7be71b3a-4714-3c76-ae94-500d5c82439a.png)
注意该方法同样会影响HttpSolr连接以及SolrCloud连接,报出下面的错误:
{"code":500,"codeMsg":"Error from server at http://xxx:8983/solr/brand: Expected mime type application/octet-stream but got text/html. <html>\n<head>\n<meta http-equiv=\"Content-Type\" content=\"text/html; charset=UTF-8\"/>\n<title>Error 401 Unauthorized</title>\n</head>\n<body><h2>HTTP ERROR 401</h2>\n<p>Problem accessing /solr/brand/select. Reason:\n<pre> Unauthorized</pre></p><hr><i><small>Powered by Jetty://</small></i><hr/>\n\n</body>\n</html>\n"}
我们可以使用简单验证的方式来测试一下:
URL url = new URL("http://xxxx:8983/solr/"); String encoding = Base64.encode("admin:xxxx".getBytes()); HttpURLConnection connection = (HttpURLConnection) url.openConnection(); connection.setRequestMethod("GET"); connection.setDoOutput(true); connection.setRequestProperty ("Authorization", "Basic " + encoding); InputStream content = (InputStream)connection.getInputStream(); BufferedReader in = new BufferedReader (new InputStreamReader(content)); String line; while ((line = in.readLine()) != null) { System.out.println(line); }
如果加上credentials,是可以成功地将结果html打印出来,否则提示401(Unauthorized)错误。
此时使用SolrCloud的情况下,就不能用CloudSolrClient中内置的httpClient,而只能在外部声明并传入,顺带建立CredentialsProvider,用于Basic权限验证:
String zkHost = PropertiesUtil.getProperty("zkhost", PROPERTY_FILE_NAME); String collection = PropertiesUtil.getProperty(collectionName.getConfigName(), PROPERTY_FILE_NAME); int zkClientTimeout = StringUtils .parseInt(PropertiesUtil.getProperty("zkClientTimeout", PROPERTY_FILE_NAME)); int zkConnectTimeout = StringUtils .parseInt(PropertiesUtil.getProperty("zkConnectTimeout", PROPERTY_FILE_NAME)); PoolingHttpClientConnectionManager connManager = new PoolingHttpClientConnectionManager(); // 连接池最大连接数 connManager.setMaxTotal(50); // 每个路由最大连接数 connManager.setDefaultMaxPerRoute(20); // http请求配置信息 RequestConfig requestConfig = RequestConfig.custom().setConnectTimeout(2000).setSocketTimeout(2000) .setConnectionRequestTimeout(500).setCookieSpec(CookieSpecs.STANDARD) .build(); CredentialsProvider credentialsProvider = new BasicCredentialsProvider(); credentialsProvider.setCredentials(new AuthScope(PropertiesUtil.getProperty("solrHost", PROPERTY_FILE_NAME), Integer.parseInt(PropertiesUtil.getProperty("solrPort", PROPERTY_FILE_NAME))), new UsernamePasswordCredentials(PropertiesUtil.getProperty("solrUserName", PROPERTY_FILE_NAME), PropertiesUtil.getProperty("solrPassword", PROPERTY_FILE_NAME))); CloseableHttpClient httpClient = HttpClients.custom().setConnectionManager(connManager).setDefaultRequestConfig(requestConfig) .setConnectionManagerShared(true).setDefaultCredentialsProvider(credentialsProvider).build(); CloudSolrClient cloudSolrClient = new CloudSolrClient(zkHost, httpClient);
此时就可以像以前一样正常查询了,注意在已经添加Credentials情况下,如果此时服务器不需要验证,也不会出错。
对定时full/delta-import的影响
如果增加了安全校验,同样也会影响full-import, delta-import的执行,此时从solr.log日志中就可以拿到错误的信息:
2016-08-29 09:10:41.925 INFO (Timer-0) [ ] o.a.s.h.d.s.BaseTimerTask [product] <index update process> Process started at .............. 29.08.2016 09:10:41 925 2016-08-29 09:10:41.948 INFO (Timer-0) [ ] o.a.s.h.d.s.BaseTimerTask [product] <index update process> Full URL http://localhost:8983/solr/product/dataimport?command=delta-import&clean=false&commit=true 2016-08-29 09:10:41.958 INFO (Timer-0) [ ] o.a.s.h.d.s.BaseTimerTask [product] <index update process> Response message Unauthorized 2016-08-29 09:10:41.958 INFO (Timer-0) [ ] o.a.s.h.d.s.BaseTimerTask [product] <index update process> Response code 401 2016-08-29 09:10:41.986 INFO (Timer-0) [ ] o.a.s.h.d.s.BaseTimerTask [product] <index update process> Disconnected from server localhost 2016-08-29 09:10:41.986 INFO (Timer-0) [ ] o.a.s.h.d.s.BaseTimerTask [product] <index update process> Process ended at ................ 29.08.2016 09:10:41 986
在full-import/delta-import的相关代码中,没有涉及到权限验证部分,直接使用http连接:
protected void sendHttpPost(String completeUrl, String coreName) { DateFormat df = new SimpleDateFormat("dd.MM.yyyy HH:mm:ss SSS"); Date startTime = new Date(); // prepare the core var String core = coreName == null ? "" : "[" + coreName + "] "; logger.info(core + "<index update process> Process started at .............. " + df.format(startTime)); try { URL url = new URL(completeUrl); HttpURLConnection conn = (HttpURLConnection) url.openConnection(); conn.setRequestMethod("POST"); conn.setRequestProperty("type", "submit"); conn.setDoOutput(true); // Send HTTP POST conn.connect(); logger.info(core + "<index update process> Full URL\t\t\t\t" + conn.getURL()); logger.info(core + "<index update process> Response message\t\t\t" + conn.getResponseMessage()); logger.info(core + "<index update process> Response code\t\t\t" + conn.getResponseCode()); // listen for change in properties file if an error occurs if (conn.getResponseCode() != 200) { reloadParams(); } conn.disconnect(); logger.info(core + "<index update process> Disconnected from server\t\t" + server); Date endTime = new Date(); logger.info(core + "<index update process> Process ended at ................ " + df.format(endTime)); } catch (MalformedURLException mue) { logger.error("Failed to assemble URL for HTTP POST", mue); } catch (IOException ioe) { logger.error( "Failed to connect to the specified URL while trying to send HTTP POST", ioe); } catch (Exception e) { logger.error("Failed to send HTTP POST", e); } }
这种方式同样需要增加权限验证, 在conn执行连接之前,要设置认证选项:
String encoding = Base64.encode(“username:password".getBytes()); conn.setRequestProperty("Authorization", "Basic " + encoding);
注意当前使用的仅仅是硬编码设置的用户名、密码,修改完成后,要将原有的jar包(apache-solr-dataimportscheduler-1.0.jar)进行重新编译并替换,重启solr服务即可。
使用内置的HttpSolrClient
即并不传入原来的,还有另外一种方法,可用于配置基本用户以及密码认证,在HttpClientConfigurer类中对DefaultHttpClient的参数进行配置时,使用了下面的属性配置:
final String basicAuthUser = config .get(HttpClientUtil.PROP_BASIC_AUTH_USER); final String basicAuthPass = config .get(HttpClientUtil.PROP_BASIC_AUTH_PASS);
如果使用这种方式,那么在创建内置的HttpSolrClient时需要传入params,在params中加入基本认证即可。
ModifiableSolrParams params = new ModifiableSolrParams(); params.set(HttpClientUtil.PROP_MAX_CONNECTIONS, 128); params.set(HttpClientUtil.PROP_MAX_CONNECTIONS_PER_HOST, 32); params.set(HttpClientUtil.PROP_FOLLOW_REDIRECTS, false); params.set(HttpClientUtil.PROP_BASIC_AUTH_USER, "admin"); params.set(HttpClientUtil.PROP_BASIC_AUTH_PASS, "zhen.com"); params.set(HttpClientUtil.PROP_MAX_CONNECTIONS, 1000); params.set(HttpClientUtil.PROP_ALLOW_COMPRESSION, true); params.set(HttpClientUtil.PROP_MAX_CONNECTIONS_PER_HOST, 1000); CloseableHttpClient closeableHttpClient = HttpClientUtil.createClient(params); HttpSolrClient solrClient = new HttpSolrClient(solrurl, closeableHttpClient);
同样的方法也适用于CloudSolrClient。