我知道 cron 作业无法通过代理运行，但是当我尝试通过此处概述的命令手动运行 security_update_db.sh 时https://help.landscape.canonical.com/LDS/ValidationGuide
sudo -u landscape bash -x /opt/canonical/landscape/scripts/update_security_db.sh
文件已下载，但当脚本经过下载部分时出现错误。
+ output=' % Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
100 11.1M 100 11.1M 0 0 355k 0 0:00:32 0:00:32 --:--:-- 315k'
+ '[' 0 -ne 0 ']'
+ mv -f /var/lib/landscape/usndb.pickle.bz2-new /var/lib/landscape/usndb.pickle.bz2
+ cd /opt/canonical/landscape
+ set -o pipefail
+ bzcat /var/lib/landscape/usndb.pickle.bz2
+ ./process-usns /dev/stdin
+ pipe_to_syslog update-security-db
+ tag=update-security-db
++ get_logger_arguments
++ echo /dev/log
++ grep -q :
++ '[' -n /dev/log ']'
++ '[' /dev/log '!=' /dev/log ']'
++ echo ''
+ args=
+ logger -s -p user.error -t update-security-db
<11>Feb 3 22:11:31 update-security-db: Traceback (most recent call last):
<11>Feb 3 22:11:31 update-security-db: File "./process-usns", line 7, in <module>
<11>Feb 3 22:11:31 update-security-db: canonical.landscape.scripts.usn.run()
<11>Feb 3 22:11:31 update-security-db: File "/opt/canonical/landscape/canonical/landscape/scripts/batch.py", line 66, in __call__
<11>Feb 3 22:11:31 update-security-db: code = self.run()
<11>Feb 3 22:11:31 update-security-db: File "/opt/canonical/landscape/canonical/landscape/scripts/usn.py", line 40, in run
<11>Feb 3 22:11:31 update-security-db: changeset = update_from_usn_tool_db(db)
<11>Feb 3 22:11:31 update-security-db: File "/opt/canonical/landscape/canonical/landscape/model/package/usn.py", line 195, in update_from_usn_tool_db
<11>Feb 3 22:11:31 update-security-db: added=added_package_usns_map, removed=removed_package_usns_map)
<11>Feb 3 22:11:31 update-security-db: File "/opt/canonical/landscape/canonical/landscape/model/package/client.py", line 38, in query
<11>Feb 3 22:11:31 update-security-db: return self._query(method, params)
<11>Feb 3 22:11:31 update-security-db: File "/opt/canonical/landscape/canonical/landscape/model/package/client.py", line 60, in _query
<11>Feb 3 22:11:31 update-security-db: raise PackageSearchRequestError(loads(error.body)["Error"])
<11>Feb 3 22:11:31 update-security-db: File "/usr/lib/python2.7/json/__init__.py", line 339, in loads
<11>Feb 3 22:11:31 update-security-db: return _default_decoder.decode(s)
<11>Feb 3 22:11:31 update-security-db: File "/usr/lib/python2.7/json/decoder.py", line 364, in decode
<11>Feb 3 22:11:31 update-security-db: obj, end = self.raw_decode(s, idx=_w(s, 0).end())
<11>Feb 3 22:11:31 update-security-db: File "/usr/lib/python2.7/json/decoder.py", line 382, in raw_decode
<11>Feb 3 22:11:31 update-security-db: raise ValueError("No JSON object could be decoded")
<11>Feb 3 22:11:31 update-security-db: ValueError: No JSON object could be decoded
+ '[' 1 -ne 0 ']'
+ alert_admin update_security_db.sh
+ echo 'Error running /opt/canonical/landscape/scripts/update_security_db.sh: 0'
Error running /opt/canonical/landscape/scripts/update_security_db.sh: 0
+ echo 'Check out the syslog output for script update_security_db.sh.'
Check out the syslog output for script update_security_db.sh.
+ exit 1
+ release_lock update_security_db.sh
+ get_distributed_lock update_security_db.sh --release
+ local command=/opt/canonical/landscape/get-distributed-lock
+ /opt/canonical/landscape/get-distributed-lock update_security_db.sh --release
+ rm -f /var/lock/update_security.lock
/var/log/landscape-server/security_update_db.log 的输出只是一遍又一遍地重复这一点：
Feb 3 16:11:38 update-security-db ERR Traceback (most recent call last):
Feb 3 16:11:38 update-security-db ERR File "./process-usns", line 7, in <module>
Feb 3 16:11:38 update-security-db ERR canonical.landscape.scripts.usn.run()
Feb 3 16:11:38 update-security-db ERR File "/opt/canonical/landscape/canonical/landscape/scripts/batch.py", line 66, in __call__
Feb 3 16:11:38 update-security-db ERR code = self.run()
Feb 3 16:11:38 update-security-db ERR File "/opt/canonical/landscape/canonical/landscape/scripts/usn.py", line 40, in run
Feb 3 16:11:38 update-security-db ERR changeset = update_from_usn_tool_db(db)
Feb 3 16:11:38 update-security-db ERR File "/opt/canonical/landscape/canonical/landscape/model/package/usn.py", line 195, in update_from_usn_tool_db
Feb 3 16:11:38 update-security-db ERR added=added_package_usns_map, removed=removed_package_usns_map)
Feb 3 16:11:38 update-security-db ERR File "/opt/canonical/landscape/canonical/landscape/model/package/client.py", line 38, in query
Feb 3 16:11:38 update-security-db ERR return self._query(method, params)
Feb 3 16:11:38 update-security-db ERR File "/opt/canonical/landscape/canonical/landscape/model/package/client.py", line 60, in _query
Feb 3 16:11:38 update-security-db ERR raise PackageSearchRequestError(loads(error.body)["Error"])
Feb 3 16:11:38 update-security-db ERR File "/usr/lib/python2.7/json/__init__.py", line 339, in loads
Feb 3 16:11:38 update-security-db ERR return _default_decoder.decode(s)
Feb 3 16:11:38 update-security-db ERR File "/usr/lib/python2.7/json/decoder.py", line 364, in decode
Feb 3 16:11:38 update-security-db ERR obj, end = self.raw_decode(s, idx=_w(s, 0).end())
Feb 3 16:11:38 update-security-db ERR File "/usr/lib/python2.7/json/decoder.py", line 382, in raw_decode
Feb 3 16:11:38 update-security-db ERR raise ValueError("No JSON object could be decoded")
Feb 3 16:11:38 update-security-db ERR ValueError: No JSON object could be decoded
我确实有一个代理设置，并配置了环境变量，将代理信息添加到 /etc/environment、/etc/apt/apt.conf、/etc/wgetrc 以及 ~/curlrc
这让我真正下载https://usn.ubuntu.com/usn-db/database.pickle.bz2但当它尝试解析时似乎失败了。有人能解决这个问题吗？
答案1
我找到了一个解决方法。脚本https://usn.ubuntu.com/usn-db/database.pickle.bz2通过代理下载后，脚本将通过管道传输到本地主机上的 Web 服务HTTP: POST http://localhost:9099/UpdateUsns HTTP/1.1
这个在代理上失败，因为代理服务器无法访问你的本地主机接口。解决方案是仅为此调用禁用代理。即
编辑
/opt/canonical/landscape/canonical/landscape/model/package/client.py
在第 56 行左右
response = self.fetch(url, post=True, data=data, total_timeout=60")
到
response = self.fetch(url, post=True, data=data, total_timeout=60, proxy="")
欢呼 / 托尔
答案2
作为不改变 LDS 源的替代解决方案，将 localhost 设置（或附加）到变量NO_PROXY中/etc/environment：
HTTP_PROXY="..."
HTTPS_PROXY="..."
NO_PROXY="localhost"
/opt/canonical/landscape/scripts/landscape-env.py所有三个变量都将由脚本update_security_db.sh（以及所有其他 cron 作业脚本）引用来获取和导出。
（在 Landscape 专用服务器 18.03 上测试。）
答案3
检查 rabbitmq 是否正常，因为当它关闭时，当它尝试处理下载的文件时也会收到 111 错误。
当 rabbitmq 运行不正常时我也遇到过同样的情况，并且rabbitmqctl status显示
Status of node rabbit@landscape
Error: unable to connect to node rabbit@landscape: nodedown
DIAGNOSTICS
===========
attempted to contact: [rabbit@landscape]
rabbit@landscape:
* connected to epmd (port 4369) on landscape
* epmd reports: node 'rabbit' not running at all
other nodes on landscape: ['rabbitmq-cli-55']
* suggestion: start the node
但是该命令rabbitmqctl start_app没有启动它并给出错误：
Starting node rabbit@landscape
Error: unable to connect to node rabbit@landscape: nodedown
我在一个论坛上找到了一份报告，报告显示清除 RabbitMQ 日志是有效的。我尝试了以下方法：
sudo service rabbitmq-server stop
sudo rm -rf /var/log/rabbitmq/*
sudo service rabbitmq-server start
之后，我可以rabbitmqctl start_app启动节点了。sudo -u landscape bash -x /opt/canonical/landscape/scripts/update_security_db.sh现在手动调用 update_security_db 脚本就可以了