及
网管TD北向接口检查维护文档
TD本地数据检查补采维护文档
一、早上和临下班前做一次TD数据检查。
1、底层入库检查
在rawdb/rawoptr@UWNMS2
查看前天至当天TD数据入库情况:正常情况下数据总数在3800左右。 观察有无明显的缺数,或者数据量偏小的情况
rawdb/rawoptr@UWNMS2
select time,TD+hwsz1 总
数,datanggz2,hwr97,hwrsz1,hwrdg1,ztergz2,ztergz3,ztergz4,zter,ericgz1,ztes,hwsz1 from
(select count(*) TD,start_time from dal_objtype_td where start_time>=trunc(sysdate - 1) group by start_time ) TD,
(select count(*) datanggz2,start_time from dal_objtype_td where omc_id=3932 and start_time>=trunc(sysdate - 1) group by start_time ) datanggz2 ,
(select count(*) hwr97,start_time from dal_objtype_td where omc_id=3803 and start_time>=trunc(sysdate - 1) group by start_time ) hwr97,
(select count(*) hwrsz1,start_time from dal_objtype_td where omc_id=3804 and start_time>=trunc(sysdate - 1) group by start_time ) hwrsz1,
(select count(*) ztergz2,start_time from dal_objtype_td where omc_id=3704 and start_time>=trunc(sysdate - 1) group by start_time ) ztergz2,
(select count(*) ztergz3,start_time from dal_objtype_td where omc_id=3705 and start_time>=trunc(sysdate - 1) group by start_time ) ztergz3,
(select count(*) ztergz4,start_time from dal_objtype_td where omc_id=3706 and start_time>=trunc(sysdate - 1) group by start_time ) ztergz4,
(select count(*) hwrdg1,start_time from dal_objtype_td where omc_id=3805 and start_time>=trunc(sysdate - 1) group by start_time ) hwrdg1,
(select count(*) zter,start_time from dal_objtype_td where omc_id=3702 and start_time>=trunc(sysdate - 1) group by start_time ) zter,
(select count(*) ericgz1,start_time from dal_objtype_td where omc_id=3103 and start_time>=trunc(sysdate - 1) group by start_time ) ericgz1,
(select count(*) ztes,start_time from dal_objtype_td where omc_id=3701 and start_time>=trunc(sysdate - 1) group by start_time ) ztes,
(select count(*) hwsz1,start_time from rawdb.dal_objtype where omc_id=3811 and start_time>=trunc(sysdate - 1) group by start_time) hwsz1, (select first_result time from npmdb.tcc_time_snap@uwnms2 where
first_result>=trunc(sysdate - 1) and first_result 2、核查2、3G融合报表-无线网-无线KPI指标报表 一早一晚核查一次报表,选择地市,小时粒度。对比查看每个地市有无明显的缺数现象 二、补采流程 1、如底层有数,报表缺数的情况 重汇总上层表数据: 1.本地数据重汇总(含补值,非补值) 登陆10.243.170.11 boco/*boco_2008$ 重汇总小时数据: RNC corba重汇(tpa_utrancell_sum表):例如话务量那些指标 perl /opt/BOCO.NPM/DB/ReportSum/bin/data_sum_q2h.pl -e '2010-09-24 02:59:59' -c frame_utrancell2rnc_15.cfg -u 0 -s '2010-09-24 02:00:00' perl /opt/BOCO.NPM/DB/ReportSum/bin/data_sum_q2h.pl -e '2010-09-24 02:59:59' -c q2h_utrancell2rnc.cfg -u -1 -s '2010-09-24 02:00:00' perl /opt/BOCO.NPM/DB/ReportSum/bin/data_sum_new.pl -e '2010-09-24 02:59:59' -c frame_utrancell_new.cfg -u 0 -s '2010-09-24 02:00:00' perl /opt/BOCO.NPM/DB/ReportSum/bin/data_sum_new.pl -e '2010-09-24 02:59:59' -c vendor_utrancell_new.cfg -u 0 -s '2010-09-24 02:00:00' RNC corba重汇(tpa_utrancell_hspa_sum表):例如HSDPA流量 perl /opt/BOCO.NPM/DB/ReportSum/bin/data_sum_q2h.pl -e '2010-09-24 02:59:59' -c frame_utrancell2rnc_hspa_15.cfg -u 0 -s '2010-09-24 02:00:00' perl /opt/BOCO.NPM/DB/ReportSum/bin/data_sum_q2h.pl -e '2010-09-24 02:59:59' -c q2h_utrancell2rnc_hspa.cfg -u -1 -s '2010-09-24 02:00:00' perl /opt/BOCO.NPM/DB/ReportSum/bin/data_sum_new.pl -e '2010-09-24 02:59:59' -c frame_utrancell_hspa_new.cfg -u 0 -s '2010-09-24 02:00:00' perl /opt/BOCO.NPM/DB/ReportSum/bin/data_sum_new.pl -e '2010-09-24 02:59:59' -c vendor_utrancell_hspa_new.cfg -u 0 -s '2010-09-24 02:00:00' RNC corba重汇(tpa_rnc_sum表):例如那些率值 perl /opt/BOCO.NPM/DB/ReportSum/bin/data_sum_q2h.pl -e '2010-09-24 02:59:59' -c q2h_rnc.cfg -u 0 -s '2010-09-24 02:00:00' perl /opt/BOCO.NPM/DB/ReportSum/bin/data_sum_new.pl -e '2010-09-24 02:59:59' -c frame_rnc_new.cfg -u 0 -s '2010-09-24 02:00:00' perl /opt/BOCO.NPM/DB/ReportSum/bin/data_sum_new.pl -e '2010-09-24 02:59:59' -c vendor_rnc_new.cfg -u 0 -s '2010-09-24 02:00:00' 上面命令参数: -t 需要触发汇总的表(一般填tpc表) -s 开始时间 -e 结束时间 重汇总日数据: 补值表: perl /opt/BOCO.NPM/DB/ReportSum/bin/data_sum_new.pl -e '2009-07-27 23:59:59' -c frame_rnc_h.cfg -u 1 -s '2009-07-27 00:00:00' 非补值表: perl /opt/BOCO.NPM/DB/ReportSum/bin/data_sum_new.pl -e '2009-07-27 23:59:59' -c frame_rnc.cfg -u 1 -s '2009-07-27 00:00:00' perl /opt/BOCO.NPM/DB/ReportSum/bin/data_sum_new.pl -e '2010-12-04 23:59:59' -c frame_utrancell_new.cfg -u 1 -s '2010-12-04 00:00:00' perl /opt/BOCO.NPM/DB/ReportSum/bin/data_sum_new_mss.pl -c frame_mscserver_v2.cfg -u 1 -s '2011-01-01 00:00:00' -e '2011-01-01 23:59:59' perl perl 23:59:59' /opt/BOCO.NPM/DB/ReportSum/bin/data_sum_new_mss.pl /opt/BOCO.NPM/DB/ReportSum/bin/data_sum_new_mss.pl -c -c serv_mscserver_v2_10026.cfg -u 1 -s '2011-01-01 00:00:00' -e '2011-01-01 23:59:59' vendor_frame_mscserver_v2.cfg -u 1 -s '2011-01-01 00:00:00' -e '2011-01-01 2.集团冻结汇总表重汇总: 登陆:10.243.170.10 boco/*boco_2008$ 集团冻结表通过10上的定时任务来跑:/opt/BOCO.NPM/DB/task/data_sum_jt_td.sh 按需要跑里面的一部分网元数据即可:例如现在需要重汇总rnc数据: 打开并复制/opt/BOCO.NPM/DB/task/data_sum_jt_td.sh中的rnc部分:(并将start_time 与stop_time替换为你需要跑的时间) perl /opt/BOCO.NPM/DB/ReportSum/bin/data_fix_td.pl -e \"$stop_time\" -c unify_rnc_h.cfg -u 0 -s \"$start_time\" perl perl /opt/BOCO.NPM/DB/ReportSum/bin/data_sum_new.pl -e \"$stop_time\" -c vendor_rnc_jt_h.cfg -u 0 -s \"$start_time\" /opt/BOCO.NPM/DB/ReportSum/bin/data_sum_new.pl -e \"$stop_time\" -c frame_rnc_jt_h.cfg -u 0 -s \"$start_time\" . 重传集团数据: 登陆10.243.170.15 pniadmin/pni321 cd /export/home/pniadmin/pni3/bin source cshrc 执行:/export/home/pniadmin/pni_perl/bin/perl sessionP_oracle.pl 其中47,48两项分别为TD的MSS与RNC的省部接口节点 所以在上传TD数据到集团的命令为: /export/home/pniadmin/pni_perl/bin/perl sessionP_oracle.pl -t 4 -b \"2009-07-25 14:00:00\" -e \"2009-07-25 14:59:59\" -d 47,48 & 上面参数含义: -t 4 手工补采选项, -b 开始时间 -e 结束时间 -d 上传数据的节点(TD为47,48) 2、底层缺数补采 1)Corba接口厂家原文件核查: 查看北向采集机的性能文件生成情况 telnet 10.170.243.1 corba/corba123 CORBA每个厂家都会以15分钟粒度送原始文件。所以一天完整数据的时间点是,一个小时4个15分钟,一天24个小时,总共96个时间目录。如有缺失不完整,则为北向采集机性能文件缺失 如果时间点完整,则需核查每个厂家的性能目录内文件,是否完整。对比正常时间点,看文件大小是否有较大出入 每个厂家的原始性能目录 厂家名称 大唐 鼎桥(退网) 华为无线 华为交换 中兴无线 中兴交换 爱立信 中兴无线(广州)1 大唐omcr2 深圳华为无线2 东莞华为无线1 中兴无线(广州)2 中兴无线(广州)3 omc_id 3931 320001 3802 3803 3811 3702 3701 3103 3704 3932 3804 3805 3705 3706 CORBA路径 /opt/NMC_CORBA/data/datang/PM / /opt/NMC_CORBA/data/dqsz1/PM /opt/NMC_CORBA/data/hwr/PM /opt/NMC_CORBA/data/hwr97/PM 数据库采集 /opt/NMC_CORBA/data/zter/PM /opt/NMC_CORBA/data/ztesgz1/PM /opt/NMC_CORBA/data/ericgz1/PM /opt/NMC_CORBA/data/ztergz2/PM /opt/NMC_CORBA/data/datanggz2/PM /opt/NMC_CORBA/data/hwrsz1/PM /opt/NMC_CORBA/data/hwrdg1/PM /opt/NMC_CORBA/data/ztergz2/PM /opt/NMC_CORBA/data/ztergz3/PM 2).CORBA接口补采(corba): 例如:中兴无线(702)缺失20090731的23时间点 1.telnet 10.243.170.1 corba/corba123 2.cd /opt/NMC_CORBA/local/bucai bucai_zter.sh 2009 07 31 23 补采2009年7月31号23点数据 注:补采程序先删除原来23点的整个目录,再重建并从厂家端ftp数据 厂家名称 大唐 新邮通 华为无线 omc_id corba补采路径 3931 3802 3803 /opt/NMC_CORBA/local/bucai/bucai_datang.sh 320101 /opt/NMC_CORBA/local/bucai/bucai_xyt.sh /opt/NMC_CORBA/local/bucai/bucai_hwr.sh /opt/NMC_CORBA/local/bucai/bucai_hwr97.sh 华为交换 中兴无线 中兴交换 爱立信 大唐omcr2 深圳华为无线2 东莞华为无线1 3811 3702 3701 3103 3932 3804 3805 数据库采集方式 /opt/NMC_CORBA/local/bucai/bucai_zter.sh /opt/NMC_CORBA/local/bucai/bucai_ztes.sh /opt/NMC_CORBA/local/bucai/bucai_ericgz1.sh /opt/NMC_CORBA/local/bucai/bucai_ztergz2.sh /opt/NMC_CORBA/local/bucai/bucai_datanggz2.sh /opt/NMC_CORBA/local/bucai/bucai_hwrsz1.sh /opt/NMC_CORBA/local/bucai/bucai_hwrdg1.sh /opt/NMC_CORBA/local/bucai/bucai_ztergz3.sh /opt/NMC_CORBA/local/bucai/bucai_ztergz4.sh 中兴无线(广州)1 3704 中兴无线(广州)2 3705 中兴无线(广州)3 3706 3).底层采集补采(npmdal) 一)大规模补采 例如:中兴无线(702)缺失20090525的1645时间点 1.telnet 10.243.170.1 npmdal/npmdal 2.cd /opt/BOCO.DAL/NPM/common/sh/tdnm/bucai bucai_zter_radio.sh 2009 07 31 00 补采2009年7月31号0点数据 bucai_zter_radio.sh里面有四个时间点00、15、30、45 如果只要补其中一个时间点,为了提供效率,可以进去把其他的时间点屏蔽 各厂家补采程序的分布情况: 爱立信 中兴交换 中兴无线(3702) 新邮通 华为无线(802) 华为无线(803) 大唐rnc 中兴无线1(3704) 深圳华为无线2 (3804) 东莞华为无线1 (3805) 中兴无线2(3705) 中兴无线3(3706) 10.243.170.2 10.243.170.1 /opt/BOCO.DAL/NPM/common/sh/tdnm/bucai/bucai_ztergz3_radio.sh /opt/BOCO.DAL/NPM/common/sh/tdnm/bucai/bucai_ztergz4_radio.sh 10.243.170.1 /opt/BOCO.DAL/NPM/common/sh/tdnm/bucai/bucai_hwrdg1_radio.sh 10.243.170.2 10.243.170.1 10.243.170.2 /opt/BOCO.DAL/NPM/common/sh/tdnm/bucai/bucai_ericgz1.sh /opt/BOCO.DAL/NPM/common/sh/tdnm/bucai/bucai_zter_radio.sh /opt/BOCO.DAL/NPM/common/sh/tdnm/bucai/bucai_hwr97_radio.sh 10.243.170.48 /opt/BOCO.DAL/NPM/common/sh/tdnm/bucai/bucai_ztes_mss.sh 10.243.170.14 /opt/BOCO.DAL/NPM/common/sh/tdnm/bucai/bucai_xinyoutong.sh 大唐carrier_cell 10.243.170.02 /opt/BOCO.DAL/NPM/common/sh/tdnm/bucai/bucai_datang_carrier_cell.sh 10.243.170.02 /opt/BOCO.DAL/NPM/common/sh/tdnm/bucai/bucai_datang_rnc.sh 10.243.170.48 /opt/BOCO.DAL/NPM/common/sh/tdnm/bucai/bucai_ztergz2_radio.sh /opt/BOCO.DAL/NPM/common/sh/tdnm/bucai/bucai_datanggz2_radio.sh /opt/BOCO.DAL/NPM/common/sh/tdnm/bucai/bucai_hwrsz1_radio.sh 10.243.170.1 大唐omcr2 (3932) 10.243.170.1 二)零星补采 对于零星的数据缺失,可以直接使用touch命令改变文件生成时间, 例如要重新采集中兴无线(3702)9月6号12点15分性能数据,并且corba数据完整 1. telnet 10.243.170.1 corba/corba123 2. cd /opt/NMC_CORBA/data/zter/PM/20090906 3. touch 1215(修改目录产生时间) 4. cd /opt/NMC_CORBA/data/zter/PM/20090906/1215 5. touch *(修改目录底下性能文件产生时间) 原理:corba接口机(.1)有程序定时将corba源文件打包给.1/.2采集机解析,打包原理是按照上一次打包结束的时间到本次要打包时间这段时间内生成的性能文件,所以修改性能文件的生成时间,能够让这个文件重新打包送去解析 4)底层采集堵塞处理 一).TD各厂家采集所在位置: 厂家名称 广州大唐无线1 深圳华为无线1 深圳中兴无线1 中兴交换 omc_id 3931 采集机ip 10.243.170.1 10.243.170.1 10.243.170.2 10.243.170.1 10.243.170.广州中兴无线1 3704 1 10.243.170.2 广州大唐无线2 深圳华为无线2 东莞华为无线1 爱立信广州无线1 广州中兴无线2 广州中兴无3932 10.243.170.2 10.243.170.2 10.243.170.2 10.243.170.1 10.243.170.1 10.243.170.采集类型 采集程序 /opt/BOCO.DAL/NPM/common/sh/tdnm/dal_datang_radio.sh 3803 /opt/BOCO.DAL/NPM/common/sh/tdnm/dal_huawei_2_cell_radio.sh 3702 /opt/BOCO.DAL/NPM/common/sh/tdnm/dal_zhongxing_radio.sh 3701 /opt/BOCO.DAL/NPM/common/sh/tdnm/dal_zhongxing_mss.sh iups_iucs_cell /opt/BOCO.DAL/NPM/common/sh/tdnm/dal_ztergz2_iups_iucs_cell.sh rnc_carrier_iub /opt/BOCO.DAL/NPM/common/sh/tdnm/dal_ztergz2_rnc_carrier_iub.sh /opt/BOCO.DAL/NPM/common/sh/tdnm/dal_datanggz2_radio.sh 3804 /opt/BOCO.DAL/NPM/common/sh/tdnm/dal_hwrsz1_radio.sh 3805 /opt/BOCO.DAL/NPM/common/sh/tdnm/dal_hwrdg1_radio.sh 3103 /opt/BOCO.DAL/NPM/common/sh/tdnm/dal_ericgz1_radio.sh 3705 3706 /opt/BOCO.DAL/NPM/common/sh/tdnm/dal_ztergz3.sh /opt/BOCO.DAL/NPM/common/sh/tdnm/dal_ztergz4.sh 线3 2 二)如果底层持续没有最新数据生成 1、登录相关omc的采集机,crontab -l 例:/opt/BOCO.DAL/NPM/common/sh/tdnm/dal_datang_radio.sh >/tmp/datang.log tail –f /tmp/datang.log 观察采集进程是否卡死 2、到厂家的采集机的采集目录下观察是不是文件过多导致程序堵塞而卡死采集进程: /opt/BOCO.DAL/NPM/data/source/tdnm/omc_id 如果是,删掉此omc_id目录内所有文件。恢复正常采集,再调用底层补采程序,补采缺数 3、过多半小时后,再重复第一步骤,看正常采集是否恢复。 三,Corba接口重启 1、telnet 10.243.170.1 corba/corba123 2、运行:ps -ef | grep java | grep ORB GDDAL4-3-/opt/NMC_CORBA> ps -ef | grep java | grep ORB corba 15962 1 0 Sep 19 ? 16:53 java tdcsmr.BcMainApp -agent ztergz4 -ORBcon corba 10648 1 0 Sep 19 ? 9:29 java tdcsmr.BcMainApp -agent zter -ORBconfig corba 5520 1 0 Sep 02 ? 22:39 java tdcsmr.BcMainApp -agent datanggz2 -ORBc corba 5514 1 0 Sep 02 ? 1:02 java tdcsmr.BcMainApp -agent datang -ORBconf corba 22266 1 0 Sep 17 ? 10:49 java tdcsmr.BcMainApp -agent hwrsz1 -ORBconf corba 5530 1 0 Sep 02 ? 8:11 java tdcsmr.BcMainApp -agent ericgz1 -ORBcon corba 7692 1 0 18:36:22 pts/2 6:48 java tdcsmr.BcMainApp -agent ztergz2 -ORBcon corba 8628 1 0 10:54:58 pts/9 0:07 java tdcsmr.BcMainApp -agent ztergz3 -ORBcon corba 4281 1 0 Sep 19 ? 7:12 java -Xbootclasspath/p:../lib/patch.jar -Xbootclasspath/p:../lib/patch.jar -Xbootclasspath/p:../lib/patch.jar -Xbootclasspath/p:../lib/patch.jar -Xbootclasspath/p:../lib/patch.jar -Xbootclasspath/p:../lib/patch.jar -Xbootclasspath/p:../lib/patch.jar -Xbootclasspath/p:../lib/patch.jar -Xbootclasspath/p:../lib/patch.jar tdcsmr.BcMainApp -agent hwrdg1 -ORBconf corba 5505 1 0 Sep 02 ? 5:14 java -Xbootclasspath/p:../lib/patch.jar tdcsmr.BcMainApp -agent ztesgz1 -ORBcon corba 5516 1 0 Sep 02 ? 17:55 java -Xbootclasspath/p:../lib/patch.jar tdcsmr.BcMainApp -agent hwr97 -ORBconfi GDDAL4-3-/opt/NMC_CORBA> kill-9 进程号 杀掉要重启OMC的进程号 3、 cdj cd bin 进入此路径: GDDAL4-3-/opt/NMC_CORBA/bin more start_runconsumer.sh /opt/NMC_CORBA/bin/runconsumer ztesgz1 /opt/NMC_CORBA/bin/runconsumer zter /opt/NMC_CORBA/bin/runconsumer datang /opt/NMC_CORBA/bin/runconsumer hwr97 /opt/NMC_CORBA/bin/runconsumer ztergz2 /opt/NMC_CORBA/bin/runconsumer datanggz2 /opt/NMC_CORBA/bin/runconsumer hwrsz1 /opt/NMC_CORBA/bin/runconsumer hwrdg1 /opt/NMC_CORBA/bin/runconsumer ztergz3 /opt/NMC_CORBA/bin/runconsumer ztergz4 /opt/NMC_CORBA/bin/runconsumer ericgz1 选择需要重启的omc 观察15分钟,告警性能有无正常上报。 配置核查 set OMC_ID=$1 perl /opt/BOCO.DAL/NRM/bin/exec.pl -o OSS$OMC_ID -f /opt/BOCO.DAL/NRM/mbin/tdnm/td_cm_init_file.sh 因篇幅问题不能全部显示,请点此查看更多更全内容