- Feature Articles
- CodeSOD
- Error'd
- Forums
-
Other Articles
- Random Article
- Other Series
- Alex's Soapbox
- Announcements
- Best of…
- Best of Email
- Best of the Sidebar
- Bring Your Own Code
- Coded Smorgasbord
- Mandatory Fun Day
- Off Topic
- Representative Line
- News Roundup
- Editor's Soapbox
- Software on the Rocks
- Souvenir Potpourri
- Sponsor Post
- Tales from the Interview
- The Daily WTF: Live
- Virtudyne
Admin
Dear Paul,
In case you can't tell, this is a grown-up place. The fact that you insist on continuously querying my IP address for NTP data clearly shows that you're too young and too stupid to be using NTP.
Go away and grow up.
Sincerely, Bert Glanstron
Admin
"Like a rock!"
Admin
Is it still too soon to say that it would fall faster than a plane piloted by JFK Jr?
Admin
Nope, not at all.
Admin
Admin
"bonzai fedoras". And there's the amusing spello that makes this submizzion complete.
"Bansai!" exclaimed Paulo. "No more work today? That's bonsa," he continued, experimenting with Auzzie slang.
Admin
Google update says "What's wrong with that?"
Admin
So, what's 'Gertrude' really like?
Admin
They used a time-based scheduler (cron) to run a command to fix the time. Does anyone find this weird?
As suggested by others, permanently running ntpd would be the best solution
Admin
Over the last year many TDWTF posts are not so much WTF, but just bugs.... plain, simple oversights or mistakes that any human could/would make. This is the latest boring "aren't i such a clever sysadmin to notice you left -e out of 'dd ibs=1 count=100M if=/dev/zero of=allzero.dat'.. hehehe" stories
Admin
Admin
I hate to poke technical holes, but for as long as I can recall, you can't run concurrent copies of ntpd because it has to bind to *:123 before it can perform any queries (ntpd uses udp/123 as its source port).
Admin
I want to hear more about this UNIVAC that ambled down its ancestors' path.
Admin
Okay, the only WTF I see here is that Bob actually had the time and inclination to contact the server owner rather than just block the offending machine from his network.
I used to contact other systems owners when we found repeated attacks coming from their machines. When this turned into something that was taking an hour plus each day to do we stopped that and instituted various tools to simply auto block the offending IPs based on a rule set for a period of time. This seems to work just fine.
Admin
I've seen that movie about 100 times and always thought the gulls said "Mate!" Googling has shown that it is officially "mine" but the producers haven't ruled it out...
Admin
Thanks, TGV. Thantgvs.
Admin
How is that a 'curious perversion in information technology', pray tell? (or interesting?)
Admin
You're supposed to have a wank on your father's underwear while looking at 3D CGIof Bea Arthur's taint while reading the article. I'm amaged you didn't know this!
Admin
I don't know why people insist on continuing to use NTP when you can get a $30 GPS receiver and have the accurate time reliably without reliance on a network connection or someone else to run a server for you.
Admin
Admin
Admin
Paul's family has a long, proud tradition of working in IT.
and Paul been the black sheep of his family due to his inexperienced and arrogance in IT.
Admin
Admin
I've seen it once, but wasn't drunk enough to think that they were saying anything besides "mine."
Admin
I don't see a WTF in this article.
Admin
http://shikan.org/bjones/Usenet.Hist/Nethist/0027.html
Unfortunately I can't find the original posting.
Admin
Actually, the problem is the local clock is too crappy that the drift cannot be fixed simply running ntpd as daemon. According to the documentation, ntpd can only fix clock drifts up to 500ppm (0.5ms drift per second, or 43.2 seconds per day). The article already said the local clock has a drift of 1-2 min per day, which obvious cannot be fixed by ntpd deamon. The only way to sync clock would be periodically running "ntpd -g -x -q". It seems that Paul in the article missed the -q option.
Admin
(forgotten to quote the original comment...)
Actually, the problem is the local clock is too crappy that the drift cannot be fixed simply running ntpd as daemon. According to the documentation, ntpd can only fix clock drifts up to 500ppm (0.5ms drift per second, or 43.2 seconds per day). The article already said the local clock has a drift of 1-2 min per day, which obvious cannot be fixed by ntpd deamon. The only way to sync clock would be periodically running "ntpd -g -x -q". It seems that Paul in the article missed the -q option.
Admin
It's the article itself.
Admin
Nice!
Admin
"That sounds like a hardware issue."
Admin
There are a couple of reasons. The first is that GPS receivers don't work in most data centers without running an external antenna. Most data centers won't or can't do that. The same applies to CDMA signals.
The second is that having a clock in tight agreement with the world is often less important than having all the clocks on your network in tight agreement with each other. The only way to do that is to have a local time server.
Admin
Modern GPS receivers are sensitive enough to work well enough completely indoors (they'll just be far less accurate at position, but that's not relevant for this application), and there's no reason why one can't be plugged into a "local time server".
Admin
Admin
Admin
Kiss of Death, not Knife of Dreams.
Admin
Had a similar experience. Our service was supposed to connect to remote database, pillage some data from it, and, if all was ok, silently terminate itself.
If all was not ok(which, for reasons outside of our control, was most of the time)... it had to run a cycle of retries(up to 15) until either all data was received or until it valiantly died trying. Except... there's been a slight oversight on my part and retry counter was being reset on each iteration...
Enter cron. Each half-hour a new instance of the service was spawned with the assumption that the previous iteration is long gone and buried.
Enter 20 hrs network disconnection period ... by the time the connection was fixed, ~40 instances of data hungry mobsters mercilessly attacked data source effectively DDOSing it.
In the end, the app was refactored to run full time with timed queries instead of being restarted each 30 minutes.
Admin
Passive-aggressive? Sorry? The owner of the DoSed machine had been very very nice.
Admin