中午收到磁盘告警:
某分区磁盘使用率超过90%
Filesystem Size Used Avail Use% Mounted on
/dev/mapper/VGSYS-lv_root
9.8G 3.0G 6.3G 32% /
tmpfs 7.8G 0 7.8G 0% /dev/shm
/dev/vda1 190M 32M 148M 18% /boot
/dev/mapper/VGSYS-lv_var
9.8G 443M 8.8G 5% /var
/dev/mapper/VGSYS-lv_ts
180G 162G 18G 91% /ts
实际上看到的/letv的空间并未占用过多空间
9.3G app
175M app-c
12K bin
191M deploy
2.6G jms-service-1
126M jms.tar.gz
36G logs
7.2M memcache
4.0K record_gc.sh
109M redis-4.0.6
123M redis.tar.gz
实际还是df 与du 的差距:
df -h 和 du -sh 出现不同大小,尤其是出现目录满的情况下,大概率会出现,df -h
满,但du -sh占用很小
1.考虑正在跑的进程,使用的文件被删除,但是空间未释放的情况
lsof |grep deleted
2.考虑inode使用的情况
如果indode 使用过多的话,一样会出现磁盘满的情况
3.考虑磁盘挂载的情况
若挂载磁盘之前,挂载点有别的文件,挂载其他设备后,会把原来的文件变为隐藏文件,同样会占用空间
首先
lsof |grep deleted
nginx 23267 worker 2w REG 253,2 873779 537338764 /ts/logs/nginx/error.log-20190312 (deleted)
nginx 23267 worker 16w REG 253,2 873779 537338764 /ts/logs/nginx/error.log-20190312 (deleted)
nginx 23267 worker 18w REG 253,2 119428723571 537239308 /ts/logs/nginx/upgrade-access.log-20190312 (deleted)
nginx 23268 worker 2w REG 253,2 873779 537338764 /ts/logs/nginx/error.log-20190312 (deleted)
nginx 23268 worker 16w REG 253,2 873779 537338764 /ts/logs/nginx/error.log-20190312 (deleted)
nginx 23268 worker 18w REG 253,2 119428723571 537239308 /ts/logs/nginx/upgrade-access.log-20190312 (deleted)
nginx 23269 worker 2w REG 253,2 873779 537338764 /ts/logs/nginx/error.log-20190312 (deleted)
nginx 23269 worker 16w REG 253,2 873779 537338764 /ts/logs/nginx/error.log-20190312 (deleted)
nginx 23269 worker 18w REG 253,2 119428723571 537239308 /ts/logs/nginx/upgrade-access.log-20190312 (deleted)
nginx 23270 worker 2w REG 253,2 873779 537338764 /ts/logs/nginx/error.log-20190312 (deleted)
nginx 23270 worker 16w REG 253,2 873779 537338764 /ts/logs/nginx/error.log-20190312 (deleted)
nginx 23270 worker 18w REG 253,2 119428723571 537239308 /ts/logs/nginx/upgrade-access.log-20190312 (deleted)
实际上是删除的错误日志仍然占用空间
重启nginx 进程之后,空间就释放了
网友评论