Yes, become the system administrator and run your server yourself. We're generally unable to help in situations where you are not the server administrator. — Michael Hampton10 secs ago
I’m administering a mail server (cPanel/WHM) running Exim. It handles normally a few hundreds incoming emails per day for our customers.
I’m not sure if I should reject SPF and/or DKIM failures for incoming email. This obviously depends on the percentage of misconfigured mail servers out there.
...
Totally covered by the canonical question. However, I'm debating adding a little blurb in there about when it's useful to report spam to the abuse@originating.domain address.
@MichaelHampton really the only time to do it is if you notice that they are hosting with a reasonably trustworthy hosting provider, and have leaked that information in the headers
like, I think reporting to the abuse addresses of consumer ISPs (especially foreign ones) is a waste of time
you know, it's really a pity that so few countries have corporeal punishment. I'd love to see people caught sending spam get stockaded and people allowed to give them the bastinado.
@FalconMomot I'm pretty sure that very few countries offer anything but corporeal punishments. The conservative churches on the other hand, handle eternal punishment rather well...
So... today's developer horror story. Drupal web site wants to integrate a forum. In order to theme it the same as the web site, they ... call curl to load the main web page at runtime, TWICE, and then insert the content of the forum page into the returned data.
@FalconMomot And the definitions are actually quite nuanced. I'd come to the false conclusion that you might be french based on the username. For which I sincerely apologize.
@MichaelHampton WTF?
Isn't it just easier to use the damn Forum module in Drupal?
So I hacked their code to load the web page, not from curl, but from Drupal's own cache. Only, now the Drupal cache is no longer caching the home page...
Fortunately I will get to yell at all the responsible parties in person next month.
@MichaelHampton Ah. Oddly, both of the "DBAs" that I worked with at $job[-2] moved on to very prestigious jobs with Drupal constulting companies. Too bad they eventually replaced those two with a clueless induhvidual.
@Adrian Well, supposedly it was temporary, until a permanent matching theme was developed to correspond with the upgrade to SMF 2.x... which has yet to happen.
@MichaelHampton Ah, yes. The old 'better ROI if we wait' routine. That's how $job[-3] ended up keeping around a single-threaded Perl-based data loader for 10 years instead of taking the 6 weeks it would need to write a C-based loader that ran 37x faster.
@MichaelHampton Beats me? Not wanting to know bad news?
I sure got the horns when I informed our common supervisor about 4 years when I tested out the disaster recovery and found that there were about 12 missing steps that existing only in the lead DBA's head.
That clueless induhvidual got promoted to be the boss. And still haven't replaced any of the Ops people that have left.
@MichaelHampton Not yet. They're still running a 6 year old Fedora Core 8 server with an 8 year old DAS SCSI160 array for their main file server. So it's going to happen in the next 18 months, I'm sure.
@MichaelHampton No shit, eh? Of course, the public-facing Fedora Core 7 web server with a connection back into their main database with cleartext usernames and passwords didn't faze them either.
Absolutely. Think of all the stuff that won't require somebody to go visit the machine anymore, for instance.
ILO means Integrated Lights Out, and that means nobody's supposed to be anywhere near the servers. That's hard to do when you have Supermicro crap that half of them don't even have IPMI cards.
@JoelESalas I think everyone is from here... never seen anything else
I think your situation was a stupid hesitation
but I'll tell a story...
we need temporary storage of ~18TB to store the data sitting on our EMC VNX units.
basically to wipe the VNXes and start over
because they're so jacked up that there's no way to fix them
so given that we sustained several EMC VNX hiccups that almost destroyed client data... what would YOU use for 18TB of NFS VMware storage... it needs to last 3-4 weeks while we rebuild the EMC units.
@ewwhite And he's probably defensive because he's backed into a corner. Seriously, how many positions are there left out there where you're still dinking with hardware? His days are numbered and he knows it.
@ewwhite explain that hobbyists do it that way, and the places you've worked where it was done properly. And if they whine, point out that they should be interested in cross-training, not bitching about it.
or if you have some other constraint... I mean, I had to manually patch super-fancy NIC card drivers in for trading systems... but getting that working was really important. THis guy was trying to get a janky solution working that would possibly fail miserably.
granted, my first internet connection was teleview, which was apparently lynx over a telnet wrapper of some sort
"A later development from Teleview provided an interfaced connection to the Internet, subscribers were given access to the Internet via a text-only terminal; email was accessed by Pine, and webpages were viewed by Lynx. Subsequently, Teleview was rendered obsolete, and SingNet started offering to the Internet via SLIP/PPP over modem."
@Iain if it's safari, make a note of the reference id at the bottom of the page. Then, when she gets kicked out check if it changed. If it does, your session was utterly lost. If it does not, some timer kicked her out.
Also, there's 2 authentication levels: one where you see yourself logged in at the top and get personal recommendations, and one where you can make bookings. We do throw people to the lower auth level on inactivity.
I created a VM with Server 2008 R2 and made it a domain controller using DC Promo but not sure what details should I put in for static IP address. Can someone assist me please ?
currently my Virtual Machine settings are,
IP 192.168.1.15 //i also tried 192.168.1.124
Default GateWay 192.16...
Inside-to-inside NAT aka NAT loopback solves hairpin NAT issues when accessing a web server on the external interface of an ASA or similar device from computers on the internal interface. This prevents DNS admins from having to maintain a duplicate internal DNS zone that has the corresponding RFC...
I don't administer our firewalls - all I know is that in order to access a website (using its public URL) thats hosted in our environment from a machine within our environment, we use a NAT loopback policy on our firewall.
@Michael I was just thinking of that quote from yesterday.
Not mentioned anywhere else, but there may be a perception that NAT loopback is a "consumer/SOHO" feature, rather than something an enterprise would use.
Interestingly, BCP 127 and BCP 142 say that routers capable of NAT MUST support hairpinning.
It's still a bad way to manage the network. Why place the onus on routers or firewalls when DNS does the job and can be controlled in a granular manner?
@ewwhite I know this is massively off-topic but it's a quiet Sunday - I'm building a new gaming rig in the next few weeks and I'm considering W8 for some reason I can't explain over 7 - I guess it's mostly in anticipation of 8.1 - which appears to be much better than 8.0 - any thoughts/considerations?
I created a VM with Server 2008 R2 and made it a domain controller using DC Promo but not sure what details should I put in for static IP address. Can someone assist me please ?
currently my Virtual Machine settings are,
IP 192.168.1.15 //i also tried 192.168.1.124
Default GateWay 192.16...
