Active prevention DWS key tool installation confirmation

1、gdb Confirm the installation （ The tool user database instance triggers core After the problem, the cluster status is abnormal repeatedly , Analyze the root causes of this problem and avoid it in time ）

Log in to any cluster node and execute the following command （HC/HCS/HCSO Environment login Outside the sandbox perform ）：

gdb --help

Prompt the following information to install

2、gstack Whether to install （ And gdb Associated tools ,gdb After installation, the tool will be installed by default , The functions and gdb identical ）

Log in to any cluster node and execute the following command （HC/HCS/HCSO Environment login Outside the sandbox perform ）：

gstack

Prompt the following information to install

gdb And gstack Please refer to the following links for installation ：

https://bbs.huaweicloud.com/forumreview/thread-182292-1-1.html

3、core Whether to configure （ This configuration ensures that the database instance triggers core The exception stack information can be retrieved after the problem , For use gdb The tool obtains the trigger instance exception from the captured information sql Timely avoidance and root cause positioning ）

The cluster status is Normal Execute the following command to confirm （ colony normal In this case, the operation does not affect the business ）

kill -11 To prepare dn Process number , Check whether the corresponding data directory generates core file , If generated core File has been configured .

If not, please follow the following link to configure ：

HC/HCS/HCSO core To configure ：https://bbs.huaweicloud.com/forum/forum.php?mod=viewthread&tid=181948

Pure soft core To configure ： https://bbs.huaweicloud.com/forum/forum.php?mod=viewthread&tid=182036

4、pg_xlogdump Whether there is （ Abnormal business produces a large number of xlog Later, the business was slow , The disk utilization rate is rising rapidly , Use this tool to resolve abnormal business ）

pg_xlogdump Prompt the following information to install （ The pure soft environment loads environment variables and then executes ,HC/HCS/HCSO Log in to In the sandbox perform ）

5、pagehack Whether there is （ The data file is damaged silently. Use this tool to parse the abnormal data file ）

pagehack Prompt the following information to install （ The pure soft environment loads environment variables and then executes ,HC/HCS/HCSO Log in to In the sandbox perform ）

pg_xlogdump And pagehack Tools get the following links ：

https://bbs.huaweicloud.com/forum/forum.php?mod=viewthread&tid=142380

The upload steps are as follows ：

step 1： Log in to the first CN node , Use omm（ Use on the cloud Ruby user ） take pagehack、pg_xlogdump Upload the tool to this node $GAUSSHOME/bin/ Next

step 2： Distribute tools to other nodes
gs_ssh -c "scp $hostname:$GAUSSHOME/bin/pagehack $GAUSSHOME/bin/"

gs_ssh -c "scp $hostname:$GAUSSHOME/bin/pg_xlogdump $GAUSSHOME/bin/"

$hostname For the first time cn Node hostname.

6、   gs_detect Tool upload steps （ This toolkit has not been developed by the O & M team , This includes the cluster status exception diagnostic tool 、IO High tools 、 Data file damage scanning and other tools , It is convenient to locate and recover problems in time ）

step 1：omm The user logs in first cn node （ Use on the cloud Ruby）, Get... In the attachment gs_detect Tool and rename it gs_detect.tar.gz Upload to the first cn node $GAUSSHOME/bin/dfx_tool/

step 2： Use the following command to unzip

cd $GAUSSHOME/bin/dfx_tool

tar -zxvf gs_detect.tar.gz

step 3： take gs_detect Distribute tools to other nodes

gs_ssh -c "scp -r hostname:$GAUSSHOME/bin/dfx_tool/gs_detect $GAUSSHOME/bin/dfx_tool/"

$hostname For the first time cn Node hostname.

  notes ： The distribution command on the cloud needs to be executed in the sandbox