So I have this Linux Virtual Machine Server at Strato and after they recently did an update I haven't been able to connect via FileZilla but get the Error
"500 OOPS: prctl PR_SET_SECCOMP failed"
I can still:
-connect via putty
-edit files
-access all pages, login, connect to the database etc
I can't:
-connect via FileZilla
What I tried
-restarted Server
-reset to a backup
-googled the error message (couldn't find anything understandable)
-contacted strato (after 20min in waiting queue "yeah, your ssh port is open, restart again, we can only check if your ssh port is open" - helpful as ever)
I have no idea what this message means or what I could do to fix it so any explanation would be greatly appreciated.
Thanks in advance,
Juliette
I have presently the same problem. Obviously, they (Strato) have changed something with their virtualization kernels. Update at Strato happened for my V-server on the 21st of Oct.
If you use vsftp then you can circumvent the FTP access problem by setting the parameter
seccomp_sandbox=NO
in your /etc/vsftp.conf.
But this is of course no permanent solution - as it reduces security.
Just a minute ago I was in contact with Strato. The guys at the helpdesk do not know whats going on. But they promised me an answer by mail.
Ralph
Related
I have a program on my computer that runs as SYSTEM and it is trying to launch an exe(opc server) on a remote machine x.x.x.x. But I get a DCOM error in my machine's eventviewer.
DCOM got error "2147942405" from the computer x.x.x.x when attempting to activate the server: yyyyy
I followed almost all the suggestions on the internet about opening dcomcnfg and adding users limits.default for launch and activation and Everyone,system,interactive,network from link (ftp://ftp.softing.com/pub/outgoing/opc/DCOM/DCOM-Settings-en.pdf) but nothing works.
Would anybody have nay other ideas on how to make this work.
According to MS Technet:
https://social.technet.microsoft.com/Forums/Azure/en-US/8bb5807f-73ba-4092-abc8-283d8fced6c4/request-a-certificate-from-certificate-service-fails-dcom-error-2147942405?forum=winserversecurity
With my VERY limited understanding of Certificate servers you may have one of the scenarios:
Client PC's are trying to connect to a Certificate server that no longer exists
Client PC's have a certificate that is valid but the Certificate server no longer exists
A Certificate server is broken
Clients do not have the proper authority to request the Certificate
I say this is limited knowledge as I am currently trying to remove AD Certificate services from a Domain Controller and I can see that in the System event log the exact same messages are being logged as I have stopped the Certificate services to asses the impact. If I get further information I will post back.
I have an issue with proFTPD.
I can see it's installed, when I look in my plesk admin (in "Updates and upgrades").
But when I try to restart proFTPD through SSH - it says "proftpd: unrecognized service". But I can see the "files" in my Plesk controlpanel. I can't connect through FileZilla.
And I don't see proFTPD in "/etc/init.d" - does anyone have and idea why this is?
Please, I really need some ideas...
I might have gone blind on this :(
proftpd running as xinetd process, you can restart it with:
/etc/init.d/xinetd restart
Error message Server sent passive reply with unroutable address points to server working behind the NAT, general advise is to switch to "active" mode as described here and pay attention to further error messages.
There maybe some other errors related to SELinux(try temporary disable it with setenforce 0) or iptables(or some intermediate firewall)
I want to send APN (Apple Push Notification) from PHP script. My PHP script is working fine and I receive notification in the iphone device as well when sending from localhost but when I upload the same script with same .pem file to the shared server it returns error...
Warning: stream_socket_client() [function.stream-socket-client]: unable to connect to ssl://gateway.push.apple.com:2195 (Connection refused) in /mypath/SendPushNotification.php on line 28 Failed to connect: 111 Connection refused
From some source I get to know that to send APN port 2195 must be open which is not possible in the shared server plan. So I tried using dedicated server and it's working perfectly and I receive the notification on the device. I tried hard to find the solution for shared server but didn't find any solution.
My question: is it possible send APN without opening port 2195 from shared server? if Yes give some guidance.
No, Its not possible to send APN message without using 2195 port.
If you want to send push notification message from shared server, try using some third party tools like http://urbanairship.com/.
Some time ago I got the same error, and then I found that it ocurres because of the server, the proxy is blocking that port, because most of servers doesn't use that port as a common one(80 or 21).
In this case you could try asking for support on your server reseller , they can open that port (2195) for you, If your take this solution I would recommend also asking to open the port 2196 that port is used for feedback.
Another solution is to use a free push notification provider, like Parse which is really easy to use and implement on your app, also you can use UrbanAirship, but I preffer Parse ;)
Hope it helps
This is irritating and definitely port 2195 issue. Chat with your hosting provider and ask them to open the port. I also had same issue. My Local code was running peacefully in every condition like sandbox and production but it was failing on the server.
Dont get confused about .pem file path and that would be simple corresponding to your code. For me i put my certificates (.pem) in a directory named certs and gave path like this:
$apnsCert = 'certs/cert_prod.pem';
The code was not working on my apps production server. I wasted one and a half day on this but that ran great on first run when i put that on application's staging server. Reason!!!! Definitely port 2195 issue.
All the beat.
There is anyway to decrypt the FTP passwords stored on Netbeans (from PHP projects) on folder called "keyring", assuming that I have informations from my windows account ???
I had lost password of one of my ftp sites.
I tried to connect with Wireshark sniffing TCP connections.
and i managed to get back the lost password.
[WRONG ! See below] No, passwords are not "encrypted" but "hashed", which is a single-way operation.
EDIT: This answer is shamefully wrong, and sits here like this since 3 years. I am deeply sorry, I don't even know why I answered something so stupid.
An easy way to get back this password would be try to connect to a server using it and dumping it using Wireshark sniffing the TCP connection.
I'm using exim on both the sending and relay hosts, the sending host seems to offer:
HELO foo_bar.example.com
Response:
501 Syntactically invalid HELO argument(s)
Possibly a problem with underscores in the hostname?
http://www.exim.org/lurker/message/20041124.113314.c44c83b2.en.html
Underscores aren't actually valid in internet host names, despite some people using them anyway. A sane DNS server should not allow you to have records for them.
Change your system's host name so it's valid, hopefully this will fix it.
After spending so many hours trying to fix this problem, which in my case just come up from nothing, I ended up with a solution. In my case, only the systems deployed to Suse OSs suddenly stopped sending emails but not those ( the same ) running on Ubuntu. After exhausting and eliminating all the suggested possibilities of this problem and even considering to change de OS of those machines, I found out that somehow the send email service is sensible to the hostname of the host machine. In the Ubuntu machines the file /etc/hosts have only the following line:
127.0.0.1 localhost
and so were the Suse machines, which stopped sending the emails. After editing the /etc/hosts from Suse machines to
127.0.0.1 localhost proplad
where proplad is the hostname of the machine, the errors were vanished. It seems that some security policy ( maybe from the smtp service ) uses the hostname information carried through the API, which was being ignored in the case of the Ubuntu machines, but not in the case of Suse machines. Hope this helps others, avoiding massive hours of research over the internet.
Diago's answer helped me solve the problem I have been trying to figure out.
Our Suse OS also stopped working out of nowhere. Tried every suggestion that I found here and on google. Nothing worked. Tried adding our domain to etc/hosts but that did not help.
Got the hostname of server with the hostname command. Added that hostname to the etc/hosts file just like Digao suggested.
127.0.0.1 localhost susetest
I saved the changes, then ran postfix stop, postfix start. And works like a charm now.
The argument to HELO should be a hostname or an IP address. foo_bar.example.com is neither an IP address nor a hostname (underscores are illegal in hostnames), so the error message is correct and there is nothing to fix.
Using qmail I came across this problem. I realised this was because of a previously unfinished installation.
1) When sending email qmail announces itself to other SMTP servers with "HELO ..." and then it adds what is in the file at: /var/qmail/control/me
(sometimes the file is located at /var/qmail/control/helohost)
2) This file should have a hostname with a valid DNS entry in.
Mine did not it had (none) which is why mails were failing to be sent.
I found that my local dev server suddenly stopped sending emails (using external SMTP) and on the server logs I found:
rejected EHLO from cpc96762-*******.net [..**.68]: syntactically invalid argument(s): 127.0.0.1:8888/app_dev.php
127.0.0.1:8888/app_dev.php is my local URL, I am using Docker, Symfony and Swift Mailer.
The only solution that helped in my case was adding the parameter:
local_domain = "localhost"
to my Swift Mailer configuration. That solved all the problems.
See the docs for the Swift Mailer local_domain option: https://symfony.com/doc/current/reference/configuration/swiftmailer.html#local-domain