We have an app which sits behind an Apache instance using mod_proxy to pass requests to the correct app server.
Our app correctly returns a Connection: Keep-Alive when queried directly, but mod_proxy is scrubbing the header.
I did some reading and a few suggested that the client needs to set a Proxy-Connection: Keep-Alive to the the proxy about this, but that didn't work either.
Is it possible to have persistent connection proxied by apache?
I have a few servers that I have personal accounts on, I have created an SVN repository to contain my home directory so that I can manage my dotfiles across these servers, which I'm finding works very well and makes it easy for me to track the changes I make (which previously would get lost/overwritten very easily).
The new problem I face is that at work I'm now forced to share one of two accounts (deploy/root) with a bunch of coworkers, and as a result I can't set too much in the way of global options.
Is there anything I can do to have the other end see my .vimrc in that session while I'm logged in (or some kludge and $VIMRC setting that makes vim load my local one?) vim is the main thing but it'd be nice to extend the principle to others as well.
I have a php appl deployed on an apache2 server running on an older release of ubuntu.
I have 4 other servers running what SHOULD be an identical configuration (evidently isn't though).
The facts I have so far:
So it looks like the issue is that php is not starting the session for ie clients, any ideas for where to start looking?
Cheers
Rich
I have a dhcp server here, running the isc-dhcpd daemon.
Some machines get an IP address fine, from syslog:
Dec 8 08:42:21 nethandler dhcpd: DHCPDISCOVER from 90:e6:ba:f6:xx:xx via eth0
Dec 8 08:42:21 nethandler dhcpd: DHCPOFFER on 10.0.0.202 to 90:e6:ba:f6:xx:xx via eth0
Dec 8 08:42:21 nethandler dhcpd: DHCPREQUEST for 10.0.0.202 (127.0.1.1) from 90:e6:ba:f6:xx:xx via eth0
Dec 8 08:42:21 nethandler dhcpd: DHCPACK on 10.0.0.202 to 90:e6:ba:f6:xx:xx via eth0
Others don't
Dec 8 08:42:04 nethandler dhcpd: DHCPDISCOVER from 08:00:27:d5:xx:xx (WIN-3053MGTDBGP) via eth0
Dec 8 08:42:04 nethandler dhcpd: DHCPOFFER on 10.0.0.132 to 08:00:27:d5:xx:xx (WIN-3053MGTDBGP) via eth0
Dec 8 08:42:08 nethandler dhcpd: DHCPDISCOVER from 08:00:27:d5:xx:xx (WIN-3053MGTDBGP) via eth0
Dec 8 08:42:08 nethandler dhcpd: DHCPOFFER on 10.0.0.132 to 08:00:27:d5:xx:xx (WIN-3053MGTDBGP) via eth0
And continue ad infinatum. I can't find any pattern to the machines that can't connect. Some wireless clients can't connect, some can. An ubuntu vm can connect, a windows 7 vm on the same hardware can. A windows server on the same hardware can't. A few macs here can't.
Any help would be appreciated, or a point to what I can use to further debug the problem.
I made an oopsie- insofar as a filesystem with important data was overwritten as an LVM volume. I never actually wrote any data to the disk though, only hosed enough of the partition table to mark it as LVM. Is there any way to recover the contents of the disk?
Anything is better than nothing.
I'm looking for an smtp service that essentially obeys the RFC, except rather than sending mail it simply logs to a file
[date] sent mail to <address>
Or whatever. I can bash this together with the bare minimum of functionality I need in python in about half an hour I reckon but if there's an existing project that works better I'd rather use that.
The reason for needing it is debugging an app that keeps sending 7* the amount of mail it's supposed to.
EDIT: And already asked: https://stackoverflow.com/questions/1006650/dummy-smtp-server-for-testing-apps-that-send-email
I'm dumping an svn repository to a giant USB disk that is formatted FAT due to necessity (treat this as unchangeable).
It conks out when you try to create a file larger than 4 gb.
I need a tool that I can pipe data to that will create files of arbitrary size that when catted together will be the original file. I can write a tool to do this, but if one already exists I'd rather use it.
Cheers
EDIT: A second look at the split man page looks like it might work.
I'm trying to deploy .Net 3.5 on an XP home SP2 machine for a customer. The install fails, with an error about WIC. Trying to install WIC on its own gave an error about not being able to verify the application, because the cryptographic service wasn't running.
A reboot got that service running, but now the install just hangs on the "inspecting your current configuration"
Task manager reports that DrWatson is doing a lot (constantly using 10% of processor time) but beyond that I don't know how to instrument what a Windows machine is doing, has anyone seen this before?
OK, this is a slightly convoluted question.
I have a machine here that collects all our mail from an upstream provider with fetchmail, and stores it in our office for collection via IMAP4.
We also have a postfix install that we use for outgoing mail in the office to shift it from the outbox quickly and let the server get around to uploading it when it feels ready.
However, my boss just told me that his wife collects her mail via POP3 directly from the upstream provider, and so his mail is not making it to her (because our postfix install is checking for a local user, not finding one and then bouncing the mail).
I can think of two solutions but I'm not familiar enough with the software to implement either
a) Configure postfix with a list of non-local addresses that it should blindly pass onto it's default relay b) Create a local account for her with a single procmail(?) rule saying to bounce the mail off to her real address, however I'd still need to specify which relay to use or it'd loop until it was blackholed.
Cheers