34 posts • joined Monday 18th June 2007 21:00 GMT
came to ask the questions to those answers. Did not leave disappoint.
When a RAID6 array busts a drive, hot spare. When a second drive heads south during the rebuild, i would immediately power down, pull each physical drive and image them, byte by byte, directly to another drive or suitable storage place. while thats going on, i replace each drive data cable in the array housing and check each drive power connecter for acceptable voltage and current. I personally feel its not a bright idea for me to let the array carry on rebuilding activities after a second disk failure within quick sucession of the first disk failure. Uptime and availability is important to the customers, but making all of their sites go offline anyway, then get months old in an instant tends to hit harder.
"A bit of downtime can be sweetened away,
not much you can do with a broken array"
P.S dont use RAID3/4/5/6 in mission critical systems, and FFS keep a good local and offsite, nightly snapshot backup with weekly and monthly rotations. Disk space is pocket change in the scheme of things.
Good point, however the transparent cache is nothing new on this particular network, what interests me is the latency being added on the first TCP req to a particular site on a particular day. A transparent cache wont do that generally (though there could be a load issue, i suppose - but i dont see how that load issue would correlate to happen only on the first access of a site on a given day). So yes, in answer to your question, I can identify the symptoms of a transparent cache, this appears to be something a little extra. My hunch is the VM DPI system, but what throws me is why the latency was not on the original docsis network, but is present on docsis 3 (again, only on the first access of a given site on a given day). This could (note careful use of word 'could' there) indicate part of virgin medias DPI and regulatory compliance strategy is to have the superhub itself do part of the logging.
I havent really had time to do a full investigation, just letting people know what I have seen.
seen an improvement
i have handsets and dongles across mutliple networks as I move around a lot and want the best chances to be contactable and online.My tmobile dongle and phone have been romaing acorss te two networks for a while now. Both get faster speeds in most places and coverage at home is much improved, which is fantastic. I usually get about 2.5mbps download and about 4mbps upload (weird huh?) from my dongle in good coverage and about 1.5mbps down and 1mbps up from my andriod. Sure, sometimes those figures drop around 6pm until 8 when most people are using things but generally i find it works well. Ive had some frustrating experiences with thier customer service in the past, but recently ive found things to work well.just signed up for another 24 months on android and my loyalty discounts bring the line rental to a paltry £4 a month so for that I cant fault them. One fault I can find though is in areas where orange has those masts that are only 2g - there is one not too far from my mothers house so when im there the phones always seem to favor the 2g orange mast, even though there is near enough line of sight to a tmobile 3g mast a few miles away.
you are probably
right on that regard. Without knowing the ins and outs of their internal network its hard to say with 100% certainty. Maybe I should have said "if it looks like a fish and swims like a fish..."
Deep packet inspection changes?
When the latest DPI was rolled out in our area (south wales) we had the same problem - some sites would not load, or only partially load (eg, some sockets worked fine, other sockets would time out).
At the time I was lucky enough to have servers in the portions of the public internet affected by the problem. I performed tcpdump packets captures on both ends of the link and noticed that some packets being sent from my servers were not arriving at the other end, despite the full SYN/ACK handshake stage taking place.
Ping and traceroutes showed zero packet loss. The problem did not exist on ports such as SSH, HTTPS or other random ports. So, to me, the problem was wholly related to something at the TCP layer on port 80. My only conclusion would be that something was meddling with the packets on that port inside of the virgin media network, which led me to the guesstimate conclusion that Virgins deep packet inspection/filtering system was going nuts.
Of course, they never admitted as such ;)
you will stay on the same
until you renew - eg change price plan. I just had a call from Tmo retentions about a similar thing. My 18 month contract is over, so im now on month to month. This does not mean that my plan 'renews' every month - it means things stay the same (3gb FUP) until I change my price plan, 'renew', or upgrade my phone (neither of which I have any intention of doing, the G1 has an excellent keyboard for SSHing).Also tmobile have a bit on their website which allows you to get a 20% discount off your monthly bill for signing a new 12 month contract, that is also classed as 'renewing' so the 500mb FUP would apply.
Easy solution: Call tmobile, say you are leaving and they give you a 20 to 30% discount without the need to sign a new 12 month contract. You stay on the month to month, are not renewing and get a better discount. Great :D
My contract ran out last month. I called retentions and sure enough - nothing they can do about the 500mb FUP, everyones got it and they seem a little pissed off about the fact ("caught them unawares" was one quote). The guy on the phone said it was because of the orange merger. He also said not many people are calling up to cancel, yet he seemed to have a very scripted response to the 'cancel over your FUP change' request, one that I could hear all the other (busy) agents around him repeating!
I told them im offski over this but am still searching for a new provider. They gave me £10 off per month until I find a new provider (and confirmed via text that this would not extend my contract in any way). It'll do till I do my research and decide what phone to get from three.
They should work great, though they would be a pain to get set up right (and do look a bit dorky to passers by as the feed appears to be pointing at the ground). I remember back in the days of consume.net fitting a cantenna to the feed of a 1m dish and scanning it around my view over the medway valley with netstumbler. Got quite a few APs (bearing in mind this is from a time when you were lucky to get one or two AP's in your street).
Alas, enough waffling. The problem with using minidishes and wifi in the EU is those pesky EIRP limits. If your AP or Wifi card and gain of antenna exceed 100mw EIRP then you are breaking the law (unless you are a ham). Though whos ever going to know .....
the problem here
Is that many of the security compromises you get on hosting platforms - malicious php code, XSS exploits and such, run as a local user - be it 'www-data' or 'nobody' (or even another user is suphp or suexec is used).
While this may not affect your average home linux user, any public facing linux server which has some vulnerable PHP code or exploitable daemon out there is now rootable. Im very much in the pro linux camp and its way orders of magnitude more secure than some other o/s's but local root exploits are very dangerous when combined with public facing servers.
Nice comment in the spolit
"/* didn't find the symbol, let's retry with the System.map dedicated to the pointlessness of Russell Coker's SELinux test machine (why does he keep upgrading the kernel if "all necessary security can be provided by SE Linux"?)*/"
Im surprised the phone even worked ;)
Wouldnt this be a shocker if the exchanges were actually pre-determined and the whole race thing is just to get names and addresses of potential future fast broadband customers.
I will be shocked *sarcasm* to see the exchanges from the 'test email' actually make it into the final.
Every time my family has notified tax credits of a change, 'glitches' seem to happen. We have disability elements in our claim.
We usually end up waiting 6 to 8 weeks a time with no tax credits for changes to our circumstances. The first two weeks of any claim is spent with the tax office denying it ever received our paperwork, until a manager finally accepts the special delivery proof of receipt as meaning they have it (pity the people who send it in regular post).
The next two weeks there are no updates, its just 'in the system' with nothing anyone can do about it..
hen week 5 they find a 'glitch' which stopped the processing during week two and which no one saw or picked up on at the time, though "someone should have." They finally sort it out in week 6 (or 7/8 depending on how they feel).
Things go well until a few months later we receive a notice that they calculated it wrong and we owe them a silly amount of money to pay back, which after a phone call they say is another 'glitch' and its really OK. A notice gets sent out saying we owe zero, then 3 months later another 'you owe us money' letter appears. Rinse and repeat all over again.
From what i read on disability forums we are on, this seems very much the norm for tax credits.
Easier way of doing it
Forge the shady XSS
honest mistake, customers benfited and they are honoring it. Just from this alone im now going to use these guys to shop. Not for the fact i may get a bargain, but for the fact that when they screw up they dont shaft their customers.
Watch and learn dell ( btw you still owe me that 2.4ghz for 50 quid from 2003!)
something just doesnt add up
If this was mossad (who are usually pretty meticulous about covering such tracks) why would they steal the identity of foreign nationals living in israel? It would just immediately leave the finger of blame pointing back to israel, which makes absolutely no sense in the 'wet work' game.
Infact, something just like this did happen to Dell in the UK - in 2003 if I recall correctly. They had a small promo of new (pretty decent) barebones base units on an overstock for 99 quid each. The URL was meant only for a certain group of customers, but word spread like wildfire and after many hours the overloaded site was shut down. I myself bought a 99 quid unit, however dell took the time to write to everyone and basicly tell us all to take a hike for taking the p*ss, even though purchase receipts has been issued. The justification was that my payment card had not been "settled" yet, only authorised. So the authorisation would lapse and no contact of sale took place. At least thats what they said ;)
Too many google hits for .cfm files on those sites. Ill bet a fiver on a new coldfusion remote exploit.
This makes me laugh. This is News is it? What I'm reading is some company decided to design an application which processes the content of a 160 byte field and returns an SMS message with a premium charge attached. Given industry standard protocols and the existence of aggregators, Id say thats not rocket science, and given the fact this has absolutely ZERO network operator support right now, how on earth did it make el reg front page?
Anam - come back when you have operator connections and agreements to make this happen. I'd be much more interested.