I couldn't find this on the interwebs and it seems like a obvious use case to me

Is it possible run Windows Desired State Configuration / DSC on a Windows 10 workstation?

For example, setting up IIS websites on development machines using DSC in "Push" mode against localhost

During performance testing of our application I found that under extreme loads one problem that can occur is the occasional mysql connection error (i.e. at this stage I don't believe it is fair to call it a bug). Eventually the application will get blocked by mysql. I can extend the SET GLOBAL max_connect_errors=10000; (currently 10) but assuming 100% mysql uptime over a very very long time would this problem eventually reoccur?

This is the "blocked-host" feature of mysql:

Host 'host_name' is blocked because of many connection errors. Unblock with 'mysqladmin flush-hosts'

Without proactive intervention will the connection error count ever flush or otherwise decline?

Update

I have found conflicting information, Percona recommend a very high limit ("around 1000000 is good") implying no. But elsewhere in the documentation states only successive errors are counted meaning yes, a single successful connection will flush the error count.

8.11.5.2. DNS Lookup Optimization and the Host Cache

If too many of these occur successively from a given host without a successful connection, the server blocks further connections from that host. The max_connect_errors system variable determines the number of permitted errors before blocking occurs

Is anyone running mysql 5.6.5 where host_cache is exposed or know definitively?

I've recently seen a lot of this IP address. A geo ip lookup tells me it is US Department of Defense Network in Arizona so why do some routers (here and here) default to this address?

The real question is should I treat 6.3.6.0 as an incorrectly reported ip address or not?

I have only anecdotal evidence of this so I wondering if am i missing some inside joke... or maybe the US Department of Defense produces routers and is particularly active :)

I'm trying use a host restricted MySQL user when connecting from one Amazon EC2 instance to another. I wanted to use elastic ip address for their dns function, but you can probably ignore the Amazon EC2 references in this question.

I think my question is how to define the dns name of the mysql client by the client when the connecting as a user to mysql?

I apologise if I have problem wrong. Perhaps it is more generic - how do define local machines host name for use with any outgoing connection?

The problem is the client app has two dns names but I need to to use the other. The details are:

Public                                    
db.fixedip.amazonaws.com                  app.fixedip.amazonaws.com
-------------------------------------------------------------------------------
Internal
db.randomdns.ec2.internal                 app.randomdns.ec2.internal

Actual machine instance
MySQL on Linux                            App on Windows 2003
'user'@'app.fixedip.amazonaws.com'

From app

mysql -uuser -p -hdb.fixedip.amazonaws.com
ERROR 1130 (HY000): Host 'app.randomdns.ec2.internal' is not allowed to connect to this MySQL server

db.fixedip.amazonaws.com conveniently resolves to the internal IP address (which can change) but the MySQL client isn't identified as its non-random app.fixedip.amazonaws.com public dns name.

This doesn't seem to work :)

mysql [email protected] -p -hdb.fixedip.amazonaws.com

UPDATE: The current answer redefines the problem - don't bother with hosts in mysql and use alternative host restriction. I'll accept that unless anyone else would like to comment on the original question

I want to use the same keypair from one availability zone in another but I cannot import it. If I generate a new key I get the same error.

See AWS console > Newtwork and Security > Key pairs > Import key pair then choose an existing EC2 keypair and I get the following error:

Value (LS0tL...tLS0t) for parameter PublicKeyMaterial is invalid. Length exceeds maximum of 1024

Why would imported public key pairs have to be smaller than the key pairs AWS generates?

I have two supposedly identical systems but one gives a 403 to POST requests from a particular tool (SOAP UI). Can I get more information about the 403 from Apache?

Background is that the server also runs SELinux, mod_jk, SSL, Tomcat, Mod_Security but their respective log files haven't helped identify the reason behind the 403 response.

I've just had two occurrences of windows 7 stalling an execution until I hit enter on the command prompt/powershell. Afterward execution seems to continue as expected.

First was an batch file of copy commands like

copy //host/file2010-1*xml localfolder/01/
copy //host/file 2010-2*xml localfolder/02/

One file seemed to be taking ages. I checked the folder and it was complete, hit enter and suddenly it burst back into life to get the rest.

Second is a java program (an ETL job) in powershell which paused as it was about to output SQL duplicate key warnings. I cannot confirm if actual execution stalled but I suspect not. I have just repeated this one as I type.

Also my more experienced colleague has seen the same thing with .net programs he has written.

How can I avoid this? What am I missing? Does Windows get lonely and need a gentle push?

I'm trying to get mod_security and hit the very common hostname error of

[alert] (EAI 2)Name or service not known: mod_unique_id: unable to find IPv4 address of "computername.domain.com"

This normally means an entry is required in the hosts file. However I have tried all combinations for the hostname; localhost (127.0.0.1), local IP address (192.168....) and public IP address ("122.56.11.156"). I have also tried all combinations for Apache's ServerName; computer.domain, domain, IP addresses and not setting it at all. Nothing seems to work for me.

Currently my /etc/hosts file looks like:

# Do not remove the following line, or various programs
# that require network functionality will fail.
127.0.0.1              computername.domain.com computername localhost.localdomain localhost    
::1             localhost6.localdomain6 localhost6
192.168.0.102           computername.domain.com computername

#192.168.0.102          computername.hostingprovider.local  computername  # NIC <eth0>
#122.56.11.156          computername.domain.com computername

Ping works, also hostname shows

[root@computername~]# hostname
computername.domain.com

The public dns doesn't have computername as a subdomain so what is the correct IP address / ServerName combo to use? What happens with duplicate entries on hosts? What does mod_unqiue_id check that could cause it to fail?

Thanks for any help you can provide

The underlying problem

Thanks for the strace suggestion. On starting apache I got the following.

1377  open("/etc/hosts", O_RDONLY)      = -1 EACCES (Permission denied)

So quickly sanity check permissions and:

# ll /etc/hosts
-rw-r--r-- 2 root root 608 Jul 22 16:01 /etc/hosts

But turned out be SELinux, see below

We are trying to find the most suitable load balancing solution that will work with our application, but I quickly found I don't know much on the subject. We were going to run Varnish Cache but we use client IP addresses at various levels and SSL - starting to get complicated... please help

Here's situation, we have one machine serving the below and looking spanning this across three. Probably share the one database.

Three types of traffic:

1. Images for emails - Basic Apache traffic (minimal)
2. Flash application - Large download(s) once per client
3. SOAP Webservice - Up to a POST per client per second

The stack:

1. Linux with PSAD
2. Apache with Mod_Security
3. Tomcat (Apache worker), JAX-WS app and MySQL database

Complications:

1. Maxmind GeoIP at app - needs client's IP address
2. mod_security on Apache - also wants IP address
3. SSL for both the Flash app and all SOAP calls

This is a preemptive measure for the client so we don't have load problems to solve as such, except if we test a DOS attack :)

Would Pound or Crossroads work? Is a software load balancer the right solution?

I don't mean to question the well established security precaution of putting apache in a chroot-jail in most scenarios but in my scenario I have doubts.

We have a virtual server hosting a single application - a webservice. This guide says I should disable SELinux which seems less secure to me? All to protect a run of the mill Centos installation after our application is compromised.

Obviously this is a bit subjective but in a small scale deployment with limited resources are there better approaches we should focus on?

What is the easiest way for me to redirect both http and https from olddomain.com to newdomain.com regardless of invalid certificate warnings?

• newdomain.com is our current domain on IIS6 and has SSL (with certificate) for some content
• olddomain.com has a permanent redirect - but this won't
  work for https://olddomain.com. It is a second website in IIS6.

I've had problems researching this because most people are concerned about wildcard SSL certificates or the complexities of multiple SSL host headers for subdomains in IIS6. Warnings seem unavoidable with SSL applied before any redirects can action. Setting up a SSL IIS6 host header just to redirect doesn't seem right (or easy) but perhaps I could change the DNS and/or use an Apache hosting account we have.

Thanks in advance

I'm having problems with third party HTML on a XAMPP based development webserver. The code produces garbage in the browser because it is filled with dubious nested quotes as follows...

<a href="$Link" title="<% sprintf(_t("READMORE","Click here to read more on &quot;%s&quot;"),$CartTitle.XML) %>">...</a>

I either need a script to bulk fix script so the code is like this (send me to Stack Overflow?):

<a href="$Link" title="<% sprintf(_t('READMORE','Click here to read more on &quot;%s&quot;'),$CartTitle.XML) %>">...</a>

OR a better understanding of the problem and a workaround because obviously other webservers handle this just fine.

This is Silverstripe (php) with asp style tags in it's templates, if that is relevant. I think it is a *nix vs windows problem but the keywords I search ("nested quotes?") produce nothing. Worse still, I know I've fixed this before! Thanks in advance.