Apache 2, running mod_chroot. Can access basic html and php script fine.
This test script works fine using cli php, so I know connection string is fine.
$host = '127.0.0.1:pubdata';
$password ="*******";
$username="SYSDBA";
$dbh = ibase_connect($host, $username, $password);
.. do query etc
However, when I run this under Chrooted Apache, it returns an error message:
Warning: ibase_connect(): Can't access lock files' directory /tmp/firebird in /html/app/webroot/test2.php
This seems really strange. /tmp/firebird exists and is used by firebird when script by CLI.
When I compiled the PHP I used --with-interbase=/opt/firebird/
Hope this makes sense and someone has a thought!
Try to use Always localhost: in front or another external IP before connecting and also restart apache + firebird before connecting
My guess is that you use Classic or SuperClassic firebird and it tries to attach to /tmp and db directly
http://firebird.1100200.n4.nabble.com/Fwd-Have-you-tried-firebird-2-5-SuperClassic-from-the-packages-td3053790.html
ps:
Also Once chrooted, Apache cannot access anything located above ChrootDir
Related
I tried asking on the Plone forums but no one had any good responses.
I am running Zope5, no ZeoServer, no Plone, with Apache as a frontend proxy.
In the old Zope2 there was a script called zodb-pack that could pack the database from the command line. This is no longer included with Zope5 and I am searching for a way to pack the db from the command line.
Also, Apache is setup for client certificate authentication, so I cannot do something like:
curl -X POST https://username:password#zope.domain.com
I also don't want to hardcode that type of curl statement because of the need to include the username and password.
My Zope is running in a Docker container, so I thought about doing something like:
source /zope5/bin/activate
python scriptname
with a python script along the lines of
from ZODB.DB import DB
from ZODB.config import databaseFromString
from transaction import commit
db = databaseFromString("<zodb_config>")
storage = db.storage
storage.pack(None, referencesf)
but I'm not sure that's the correct way to do this. Basically I just want my bash script that automates the backups for the server to pack the Zope DB before backing it up, but I need a command line command to do so.
I cannot use any solution that requires me to modify how Zope runs, nor requires me to stop Zope to perform the pack.
Of course I can manually go to the ZMI's Control Panel and click Pack, but like I said, I was trying to automate it so it could run in off peak hours.
I try to install Artifactory 7.11 on linux instance.
Detiles:
Cloud: AWS
OS: amazon linux
If i start the service with the default configuration everything work properly.
But if i uncomment the database section in the system.yaml file to work with SQL server i got an error that the "system.yaml validation failed" at node .shared.database.
Note that the connection details like user,password are checked and worked.
Additionally - the sql port(1433) port is already opened.
As you can see that the connection to the DB succeeded:
Default configuration file looks like this:
Configuration file with connection to the DB
Does anyone have an idea what can be the issue?
Thanks.
After a long examination I finally understood the main reason...
default the IPv6 is configured, so i've disable the ip section in the system.yaml and put the IPv4 server ip
multiple sql drivers was in the folder so i've deleted all of them except the v11.jar
Thanks #Prasanna for your help!
SETUP
I have a windows 10 PC connecting to my linux server (ubuntu 18.04.2). I run/edit code on my powerful server by remotely accessing the server through VS code's ssh ability. It is super convenient to edit my code on the server with VS code...
THE ISSUE
Visual Studio code won't connect to my Linux server. What is weird is that I am able to ssh into the server from the terminal within VScode by just literally typing ssh <username>#<server-hostname>
However, when I do the Remote-SSH option within VScode then it does not want to connect to the server...strange...
Here is how I remotely SSH into server with VScode built in functionality (marked with red rectangle). Also part of this screenshot is my config file, which I have triple checked as correct (IP address and .ssh/id_rsa_gpu_1_solarpv_ssh location)
HERE IS THE ERROR OUTPUT
MY ATTEMPTS
I have made sure that the config file is setup correctly and that the IP address is correct.
I have ensured that the server's IP address is also in the 'known_hosts' file located at /Users/.ssh/
I even generated my own private key for SSH connection as advised by the great documentation here: https://code.visualstudio.com/docs/remote/troubleshooting#_quick-start-using-ssh-keys
Still getting the error of failed to create hard link...
With all of my back-end stuff done right, I am absolutely clueless regarding why this is an issue.
After spending literally 9 hours on this issue...here is the answer:
When you launch VScode and remotely access the linux server, the linux server is instructed to download and install a vscode server so that you (on the other side) can remotely connect to the server.
If the server does not have internet just make sure the following setting is ticked
This allows the remote server (without or with internet connection) to get the vscode server running.
To eliminate the "failed to create hard link error" go to the directory mentioned by the error, which in this case on linux server is
navigate to /home/<username>/.vscode-server/bin/0ba0ca.../
then remove/delete the hardlink file and target file which also has
the same name as the directory.
Then just attempt to remotely access your server again through
vscode and all should be fine
Here is also a discussion regarding the issue...if you still have problems https://github.com/microsoft/vscode-remote-release/issues/2507
I solved this by enabling the option Lockfiles in Tmp:
i just deleted the whole library and it works, i don't see anything wrong with doing that.
rm -r /path_to_project/.vscode-server/
if i'm wrong feel free to correct me :)
Try this
Just in case others need to try the same here is the detail
ctrl-shift-p and choose:
Remote-SSH: kill VS Code Server on Host...
Chose the host name
Selected the platform type (linux)
https://github.com/microsoft/vscode-remote-release/issues/4307#issuecomment-762882247
worked for me.
I'm trying to connect to an admin server in WLST using config and key files. There are no error messages but I am prompted for a username and password. These files were created (by another developer who is long gone[1]) with the storeUserConfig() command. My call to connect looks something like this: connect(userConfigFile=configFile, userKeyFile=keyFile, url='t3://somehost:7031')).
Is there some restriction in using these files, such as it can only be used on the host where created, or it needs access to the domain's boot.properties file?
Note: I'm trying to connect to an admin server on a different host and non-standard port (e.g. not 7001). The server I am running WLST on and the remote host are the same version of Weblogic.
Some of the things I have tried:
verified that these files appear correct, the key file being binary data and the config file having a line for "weblogic.management.username={AES}..." and "weblogic.management.password={AES}...".
verified that there is a server on the specified port by entering a known login and password that is successful
specified the admin server in the connect parameter
turn on debug(true); the only output is <wlst-debug> connect : Will check if userConfig and userKeyFile should be used to connect to the server and another line giving the path to the userConfig file
turn on Python logging in jython with -Dpython.verbose=debug; nothing relevant to decryption operation
Munging the key or the config files generates no error messages and behaviour as above
[1]: These files are still used today by other existing WLST scripts. However, these scripts are so convoluted and deliberately obfuscated that they are very difficult to reverse-engineer how connect() is being called.
You do not need to access to the domain's boot.properties file. You just need to make sure the configFile and keyFile pointing to the right files. FYI, here is one of the commands we are using:connect(userConfigFile='./user.secure',userKeyFile='./key.secure',url='t3://somehost:7001')
Have you check the network connectity that might be having a firewall in between that troubling you, check the traceroute from the script machine to the Remote machine. Recently I have faced simalar issue. once the routing table updated with allow the WL admin server port everything got set.
Hope this could helps you!
I had this problem too. In a script, I exported the Linux variables userConfigFile and userKeyFile. Then I connected by running:
url='t3://localhost:7002'
userConfigFile='$userConfigFile'
userKeyFile='$userKeyFile'
connect(userConfigFile=$userConfigFile, userKeyFile=#userKeyFile, url=url)
That all worked in a script, but would not work interactively. I changed to doing the following:
url='t3://localhost:7002'
userConfigFile='/users/me/weblogic-2014/weblogic-admin-WebLogicConfig.properties'
userKeyFile='/users/me/weblogic-2014/weblogic-admin-WebLogicKey.properties'
connect(userConfigFile=userConfigFile, userKeyFile=userKeyFile, url=url)
And that worked interactively.
Has anyone find to use the GAE remote api but instead of connecting to AppEngine to connect to localhost?
For dev purposes of course
i was able to get this working by adding the following to the app.yaml file
builtins:
- remote_api: on
and then from the command line you can access the db, users, urlfetch or memcache modules
remote_api_shell.py -s localhost:8080
This will prompt you for the email and password but this is not important right now. the remote_api_shell.py is on my path from the google app engine directory
Have you tried the development console? To access it, go to this URL: http://localhost:8080/_ah/admin.
If you really want to use the remote API, have a look at this article. I believe you can use the dev_server by passing the local host url to the interactive console script.
For Java see this document which explains both local and remote access
https://developers.google.com/appengine/docs/java/tools/remoteapi#Configuring_Remote_API_on_the_Client
If there are some like me who prefer to execute from a python script rather than a shell:
from google.appengine.ext.remote_api import remote_api_stub
remote_api_stub.ConfigureRemoteApiForOAuth('localhost:8081', '/_ah/remote_api', secure=False)
os.environ['SERVER_SOFTWARE'] = 'Development'
os.environ['HTTP_HOST'] = 'localhost:8080'
... do stuff ...
I run the dev server with the option "--api_port 8081" otherwise just look at the port used in the dev server logs ("Starting API server at ...").
The environ tweaks are to be able to use cloudstorage api against the dev server too.