S.L. Barth's questions

I needed to do some work on a KVM guest. To be able to go back to it's previous state, I took a snapshot:

virsh snapshot-create vph-share

After this, I tried to SSH into the VM:

ssh vph-share.hexvm

But I got

ssh: connect to host vph-share.hexvm port 22: No route to host

This worked fine before I took the snapshot.

I took a look at the snapshot-list, it looked normal. I then tried to log in again, and got

ssh: connect to host vph-share.hexvm port 22: Connection timed out

I also tried a virsh shutdown, but although virsh reports that the VM is being shutdown in reality it isn't. A following virsh start says that the VM is already active.

I tried virsh reboot and got:

error: Failed to reboot domain vph-share
error: Requested operation is not valid: Reboot is not supported without the JSON monitor

I tried the same actions on another VM on my server, without problems.

The log (/var/log/libvirt/qemu/vph-share.log) only has a note about the shutdown attempt:

2015-03-19 09:14:43.445+0000: shutting down

How can I get this VM back to work again?

The OS of the host machine is Ubuntu Server 12.04.
The OS of the KVM guest is Ubuntu 12.04.

I am migrating a mailman installation from an Ubuntu 8.0.4 server to an Ubuntu 12.04 server.

I have succesfully installed mailman on the target server and copied the data, archives and lists directories there as per this article.

Then I tried sudo service mailman start and was greeted with

KeyError: 'ee'
 * Site list for mailman missing (looking for list named 'mailman').
 * Please create it; until then, mailman will refuse to start.

When I tried to create the mailman list with bin\newlist mailman (as per the official documentation) it tells me that the list mailman already exists. Trying to start or restart fails.

So I changed the mm_cfg.py file, as per the documentation again, and set

real_name = 'Mailman2'

Giving the command

/var/lib/mailman$ sudo bin/config_list -i data/sitelist.cfg mailman2

yields:

Non-standard property restored: personalize

which sounds like things worked. However when I try sudo service mailman start, I get the same warning for the list mailman2:

KeyError: 'ee'
 * Site list for mailman missing (looking for list named 'mailman2').
 * Please create it; until then, mailman will refuse to start.

How can I create a "site list" for mailman and start the mailman service?

I have been asked to disable a single mailing list on mailman. I do not want to destroy the list - we may want to revive it later. Also, we want to keep the archives accessible for the list members.

We are using mailman 2.1.14, with Postfix 2.9.6, on a VM running Ubuntu 12.04.

It seems that mailman does not have an option to disable a list. Searching the web, I found that I should instead solve this at the level of the MTA.
Specifically for Postfix, I found that I should use header_checks.
(For reference, I found a similar advice for sendmail users).

I have updated the /etc/postfix/main_cf:

# Disable mailing lists
header_checks = pcre:/etc/postfix/header_checks.pcre

And created an /etc/postfix/header_checks.pcre file containing:

/^Sender: <[email protected]>/ REJECT
/^Sender: <[email protected]>/ REJECT

(All mails come with "-bounces" attached to the Sender, for some reason).

Then I sent a mail to the test mailing list, and it did not arrive or appear in the archives - as intended.
A mail I had sent before changing the configuration had arrived in my mailbox and in the archives, also as intended.

The question is - is this a proper way of doing this? Or could this setup cause me trouble later on?

I use Protected Configuration to store sensitive data encrypted in the Web.config file of our asp.net application.
I have followed the steps in Walkthrough: Creating and Exporting an RSA Key Container to generate and export a private key.
I have followed the steps in Walkthrough: Encrypting Configuration Information Using Protected Configuration to encrypt a section of the Web.config file using that key.

My problem is that I am unable to use that key on the test environment, due to strange behaviour from the aspnet_regiis command.

I have been able to get the thing to work on a colleague's computer that runs Windows 8. Both the aspnet_regiis -pi command (importing the private key) and the aspnet_regiis -pa (granting access to the private key) worked without problems there.

On the test environment that runs Windows 7, I have been able to successfully import the private key into the key container using aspnet_regiis -pi <KeyContainer> <KeyFile>.
But when I try aspnet_regiis -pa <KeyContainer> <account-name>, the command just lists it's command line options. It does not indicate failure or success.

All these environments use .Net Framework v4.

I have looked at this question, but the difference is that I was able to execute aspnet_regiis -pi <KeyContainer> <KeyFile>. Nor did it make a difference to execute aspnet_regiis -lk as is suggested there.

The question is: how can I grant the application access to the contents of the Key Container, so that it can decrypt the Web.config? Why is aspnet_regiis acting as if the "-pa" option does not exist, rather than giving me a "Success" or "Failed" message?