I received a monitoring email from Baidu in the morning and found that the website could not be opened, and there was nothing to worry about at that time, because the website could not be opened, and the possible reasons were as follows:
1: The server is in arrears
2: The server has been attacked
At that time, I didn't pay too much attention, brushed my teeth and washed my face to go to work, and then turned on the computer to visit the Alibaba Cloud website, and then to see if it was attacked or overdue, I found that there was no arrears or attack.
Then, ssh connected to the server, restarted the server, waited for 5 minutes, and found that the server could not be connected, and found that things were not that simple!!!
Since the external network link cannot be connected, let's use VNC to connect, that is, log in to the Alibaba Cloud website console to have a remote connection, which is the web version of the remote connection, and the intranet should be walked, and then, find, black screen!!!
Things are getting less and less simple, I feel like there is something wrong with Alibaba Cloud??? (Because, I haven't touched any operation on the server) First, I submitted a work order, explained the situation, and then went to work...
On the subway, log in to the Alibaba Cloud APP, pay attention to the work order feedback at any time, suddenly, on the app page, there is an announcement, link: https://help.aliyun.com/noticelist/articleid/20651342.html
(Restored) On November 12, 2017, Beijing time, the telecom north-south mutual visit was announced
Personally, I guess it is a problem on Alibaba Cloud's side, and whether it is caused by some official Alibaba Cloud services or interfaces that have not returned to normal????
After the Alibaba Cloud ticket feedback, I found that I could use ssh to the server, but after restarting, I couldn't connect to the server, and I repeatedly communicated with the ticket, and the phone was communicated for a long time.
Even if SSH can be connected, some services of the server cannot work normally, such as mysql server, which does not start normally...
Finally, the ticket replies as follows:
Host remote connection as well as recovery, you test down.
Current problem: There is an abnormal startup item set in rc.local, causing the system to fail to fully boot after starting, and the card is mastered, as shown in Annex 1.
After commenting out the content in rc.local, it will return to normal, please check whether there are any abnormalities in the service.
After processing, open the content in rc.local.
I seem to understand something, it is that some services of the server did not start normally, and then I manually started mysql and found that it could not be started in 30 seconds, and I found that things were not that simple。。。。。
Through my personal intuition, I hit the df -h command to check the hard disk space usage, and found that there was a piece of space that took up 100% of the space!!!!
Finally, find the top 10 files that take up disk space, and find a file, more than 12g, which is a log file of nginx,
Should beDisk space is full, resulting in a failureModify the configuration parameters of nginx logs, then delete the nginx log files, restart the server, and some will return to normal.
nginx configuration has a pit.When the server segment does not specify a access_log, and no access_log parameters are specified in the http segment, it will be written to the logs/access.log file by default, that is, the default value access_log "logs/access.log", and it is the access log of all servers。
|
Posted on 11/13/2017 12:16:27 PM
|
|
|
|
Landlord|
Posted on 11/13/2017 12:22:36 PM
|
