#作者:程宏斌
文章目录
-
- [fluent-bit 1.9.4 转换测试](#fluent-bit 1.9.4 转换测试)
- 结论
接上篇:《Fluent Bit系列:字符集转码测试(上)》https://blog.csdn.net/qq_40477248/article/details/150776142?spm=1001.2014.3001.5501
fluent-bit 1.9.4 转换测试
1、测试使用配置文件
[SERVICE]
Flush 1
Parsers_File parsers.conf
HTTP_Server On
HTTP_Listen 0.0.0.0
HTTP_PORT 3194
[INPUT]
Name tail
Tag regex-fluent
DB ./db/regex-fluent.db
Read_from_Head true
Path /var/log/pods/logtest/*.log
Path_Key pod_log_path
[FILTER]
Name modify
Match *
Add paas_log_belong user
Add paas_log_type middleware
Add paas_collection_type userfile
Add paas_account_id 123456789
Add paas_region_id lftst
Add paas_product_id ccc
Add paas_instance_name test10
Add paas_host_ip 127.0.0.1
Add paas_manager_ip 127.0.0.1
Add pod_namespace default
Add pod_name test-0
Add pod_container_name test
[FILTER]
Name multiline
Match *
multiline.key_content log
multiline.parser multiline-regex-go
emitter_mem_buf_limit 2048M
[FILTER]
Name lua
Match *
script gbk2utf8.lua
call convert_gbk_to_utf8
[OUTPUT]
Name file
Match *
Path /vdata/logtest
核查数据结果
# 检查采集结果命令
less /vdata/logtest/regex-fluent #提前获取中文字符
egrep -n '匹配费率为0,不进行计价|to_number' regex-fluent | head -n 5
以下是采集后的结果被输入到 /vdata/logtest/regex-fluent 文件中。在Linux系统中,通过使用 grep 命令搜索关键字,可以看到文件的第135、137行,确认没有出现任何 GBK 格式的乱码。这表明 Fluent Bit 1.9.4 成功地将 GBK 格式转换为 UTF-8。此外,第24、40、41行的内容也确认了日志被正确合并。 但是我发现这个正则没有处理日志文件中的空行,导致采集的数据有空白。
[root@cdp-10-191-193-8 logtest]# egrep -n '匹配费率为0,不进行计价|to_number' regex-fluent | head -n 5
24:regex-fluent: [1720149989.173038023, {"pod_log_path":"/var/log/pods/logtest/20230911_filerate-3105-9c659b484-kfp5j.0.log","paas_host_ip":"127.0.0.1","paas_log_belong":"user","pod_namespace":"default","paas_log_type":"middleware","pod_name":"test-0","paas_collection_type":"userfile","pod_container_name":"test","paas_account_id":"123456789","paas_region_id":"lftst","paas_manager_ip":"127.0.0.1","paas_product_id":"ccc","log":"329 2023-09-11 17:03:30 [/data01/heht30/guoc//stl_jsdev/src/load_shm.c:fLoadTBillCycle:401] [INFO]sql is [select \na.bill_cycle_seq, \na.bill_period_id, \nto_number(to_char(nvl(a.start_date, add_months(sysdate, 0)), 'yyyymmdd')), \nto_number(to_char(nvl(a.cutoff_date, add_months(sysdate, 240)), 'yyyymmdd')), \nnvl(a.split_table_postfix,-1), \nnvl(b.latn_id,-1), \nto_number(b.status) ","paas_instance_name":"test10"}]
40:regex-fluent: [1720149989.173041374, {"pod_log_path":"/var/log/pods/logtest/20230911_filerate-3105-9c659b484-kfp5j.0.log","paas_host_ip":"127.0.0.1","paas_log_belong":"user","pod_namespace":"default","paas_log_type":"middleware","pod_name":"test-0","paas_collection_type":"userfile","pod_container_name":"test","paas_account_id":"123456789","paas_region_id":"lftst","paas_manager_ip":"127.0.0.1","paas_product_id":"ccc","log":" to_number(nvl(to_char(active_date,'yyyymmdd'),'19700101')), ","paas_instance_name":"test10"}]
41:regex-fluent: [1720149989.173046930, {"pod_log_path":"/var/log/pods/logtest/20230911_filerate-3105-9c659b484-kfp5j.0.log","paas_host_ip":"127.0.0.1","paas_log_belong":"user","pod_namespace":"default","paas_log_type":"middleware","pod_name":"test-0","paas_collection_type":"userfile","pod_container_name":"test","paas_account_id":"123456789","paas_region_id":"lftst","paas_manager_ip":"127.0.0.1","paas_product_id":"ccc","log":" to_number(to_char(nvl(inactive_date,add_months(sysdate,360)),'yyyymmdd'))","paas_instance_name":"test10"}]
135:regex-fluent: [1720149989.173070565, {"pod_log_path":"/var/log/pods/logtest/20230911_filerate-3105-9c659b484-kfp5j.0.log","paas_host_ip":"127.0.0.1","paas_log_belong":"user","pod_namespace":"default","paas_log_type":"middleware","pod_name":"test-0","paas_collection_type":"userfile","pod_container_name":"test","paas_account_id":"123456789","paas_region_id":"lftst","paas_manager_ip":"127.0.0.1","paas_product_id":"ccc","log":"329 2023-09-11 17:04:53 [/data01/heht30/guoc//stl_jsdev/src/stl_pub.c:CdrRating:761] [INFO]匹配费率为0,不进行计价","paas_instance_name":"test10"}]
137:regex-fluent: [1720149989.173070837, {"pod_log_path":"/var/log/pods/logtest/20230911_filerate-3105-9c659b484-kfp5j.0.log","paas_host_ip":"127.0.0.1","paas_log_belong":"user","pod_namespace":"default","paas_log_type":"middleware","pod_name":"test-0","paas_collection_type":"userfile","pod_container_name":"test","paas_account_id":"123456789","paas_region_id":"lftst","paas_manager_ip":"127.0.0.1","paas_product_id":"ccc","log":"329 2023-09-11 17:04:53 [/data01/heht30/guoc//stl_jsdev/src/stl_pub.c:CdrRating:761] [INFO]匹配费率为0,不进行计价","paas_instance_name":"test10"}]
下面是日志文件的前三行
[root@cdp-10-191-193-8 logtest]# head -n 3 /home/dcos/fluentbit/20230911_filerate-3105-9c659b484-kfp5j.0.log
329 2023-09-11 17:03:29 [/data01/heht30/guoc//stl_jsdev/src/stl_main.c:main:328] [INFO][REPEAT_MESSAGE=2|MESSAGE_SEQUENCE=3105:72:239|THIS_WORKFLOW_ID=72|INSERT_TIME=20230911164216|THIS_NODE_ID=3|GROUP_ID=3105|WORKFLOW_ID=72|FILE_NAME=991000PTSVDA022023090116581100000005.filefmt|PROVINCE_ID=99|FILE_PATH=/jzjs_month/spcp//2799/20230901/99|FMT_NORMAL_REC=10853], ���ۿ�ʼʱ��=[2023-09-11 17:03:29]
329 2023-09-11 17:03:29 [/data01/heht30/guoc//stl_jsdev/src/stl_main.c:main:341] [ALERT]REPEAT_MESSAGE=2|MESSAGE_SEQUENCE=3105:72:239|THIS_WORKFLOW_ID=72|INSERT_TIME=20230911164216|THIS_NODE_ID=3|GROUP_ID=3105|WORKFLOW_ID=72|FILE_NAME=991000PTSVDA022023090116581100000005.filefmt|PROVINCE_ID=99|FILE_PATH=/jzjs_month/spcp//2799/20230901/99|FMT_NORMAL_REC=10853
下面是采集文件的前三行内容,第二行也就是时间戳为"1720149989.173032110"的行,没有正常处理空白行。
[root@cdp-10-191-193-8 logtest]# head -n 3 /vdata/logtest/regex-fluent
regex-fluent: [1720149989.173026110, {"pod_log_path":"/var/log/pods/logtest/20230911_filerate-3105-9c659b484-kfp5j.0.log","paas_host_ip":"127.0.0.1","paas_log_belong":"user","pod_namespace":"default","paas_log_type":"middleware","pod_name":"test-0","paas_collection_type":"userfile","pod_container_name":"test","paas_account_id":"123456789","paas_region_id":"lftst","paas_manager_ip":"127.0.0.1","paas_product_id":"ccc","log":"329 2023-09-11 17:03:29 [/data01/heht30/guoc//stl_jsdev/src/stl_main.c:main:328] [INFO][REPEAT_MESSAGE=2|MESSAGE_SEQUENCE=3105:72:239|THIS_WORKFLOW_ID=72|INSERT_TIME=20230911164216|THIS_NODE_ID=3|GROUP_ID=3105|WORKFLOW_ID=72|FILE_NAME=991000PTSVDA022023090116581100000005.filefmt|PROVINCE_ID=99|FILE_PATH=/jzjs_month/spcp//2799/20230901/99|FMT_NORMAL_REC=10853], 批价开始时间=[2023-09-11 17:03:29]","paas_instance_name":"test10"}]
regex-fluent: [1720149989.173032110, {"pod_log_path":"/var/log/pods/logtest/20230911_filerate-3105-9c659b484-kfp5j.0.log","paas_host_ip":"127.0.0.1","paas_log_belong":"user","pod_namespace":"default","paas_log_type":"middleware","pod_name":"test-0","paas_collection_type":"userfile","pod_container_name":"test","paas_account_id":"123456789","paas_region_id":"lftst","paas_manager_ip":"127.0.0.1","paas_product_id":"ccc","log":" ","paas_instance_name":"test10"}]
regex-fluent: [1720149989.173032547, {"pod_log_path":"/var/log/pods/logtest/20230911_filerate-3105-9c659b484-kfp5j.0.log","paas_host_ip":"127.0.0.1","paas_log_belong":"user","pod_namespace":"default","paas_log_type":"middleware","pod_name":"test-0","paas_collection_type":"userfile","pod_container_name":"test","paas_account_id":"123456789","paas_region_id":"lftst","paas_manager_ip":"127.0.0.1","paas_product_id":"ccc","log":"329 2023-09-11 17:03:29 [/data01/heht30/guoc//stl_jsdev/src/stl_main.c:main:341] [ALERT]REPEAT_MESSAGE=2|MESSAGE_SEQUENCE=3105:72:239|THIS_WORKFLOW_ID=72|INSERT_TIME=20230911164216|THIS_NODE_ID=3|GROUP_ID=3105|WORKFLOW_ID=72|FILE_NAME=991000PTSVDA022023090116581100000005.filefmt|PROVINCE_ID=99|FILE_PATH=/jzjs_month/spcp//2799/20230901/99|FMT_NORMAL_REC=10853 ","paas_instance_name":"test10"}]
结论
Fluent Bit 1.9.4 和 3.0.2 均能够通过此 Lua 脚本进行字符集转换验证,证明字符集转码准确性验证成功。
由于旧版本正则表达式对空白行不进行处理,接下来所有字符集测试都基于新版本正则表达式。因此,不再使用旧版本和新版本正则表达式进行性能对比测试,因为旧版本正则表达式功能不完善,不再具备测试的必要性。