In October last year I noted that the Western Digital “Green” drives in my desktop and a new RAID server build looked to be in imminent danger of early failure. That conclusion was based on a worryingly high load-cycle count which a series of posts around the net all attributed to the aggressive head parking features of these drives in order to save energy when not in use. I decided at the time to replace the desktop disk and recycle it to the RAID server. I have since decided to replace the entire RAID array as soon as I can. But which disks to use?
Well, Backblaze, a company which offers “unlimited” on-line backup storage for Mac and Windows users, has just published a rather useful set of statistics on the disks that they use in their storage arrays. The most interesting point is that they use the same domestic grade disks as would be used by you or I rather than the commercial grade ones used in high end RAID systems.
According to the backblaze blog post, at the end of 2013 they had 27,134 consumer-grade drives spinning in their “Storage Pods”. Those disks were mostly Seagate and Hitachi drives, with a (much) lower number of Western Digital also in use. Backblaze said:
Why do we have the drives we have? Basically, we buy the least expensive drives that will work. When a new drive comes on the market that looks like it would work, and the price is good, we test a pod full and see how they perform. The new drives go through initial setup tests, a stress test, and then a couple weeks in production. (A couple of weeks is enough to fill the pod with data.) If things still look good, that drive goes on the buy list. When the price is right, we buy it.
They also noted:
The drives that just don’t work in our environment are Western Digital Green 3TB drives and Seagate LP (low power) 2TB drives. Both of these drives start accumulating errors as soon as they are put into production. We think this is related to vibration. The drives do somewhat better in the new low-vibration Backblaze Storage Pod, but still not well enough.
These drives are designed to be energy-efficient, and spin down aggressively when not in use. In the Backblaze environment, they spin down frequently, and then spin right back up. We think that this causes a lot of wear on the drive.
Apart from the vibration point, I’d say that conclusion was spot-on given the reporting I have seen elsewhere. And the sheer number of drives they have in use gives a good solid statistical base upon which to draw when making future purchasing decisions. Backblaze note that the most reliable drives appear to those made by Hitachi (they get “four nines” of untroubled operation time, while the other brands just get “two nines”) but they also note that the Hitachi drive business was bought by Western Digital around 18 months ago – and WD disks do not seem to perform anywhere near as well as the others in use.
The post concludes:
We are focusing on 4TB drives for new pods. For these, our current favorite is the Seagate Desktop HDD.15 (ST4000DM000). We’ll have to keep an eye on them, though. Historically, Seagate drives have performed well at first, and then had higher failure rates later.
Our other favorite is the Western Digital 3TB Red (WD30EFRX).
We still have to buy smaller drives as replacements for older pods where drives fail. The drives we absolutely won’t buy are Western Digital 3TB Green drives and Seagate 2TB LP drives.
So I guess I’ll be buying more Seagates in future – and I was right to dump the WD caviar green when I did.
(As an aside, I’m not convinced the Backblaze backup model is a good idea, but that is not the point here).
Just a quick note to say our January meeting will be on Wednesday night at 7.30pm, usual place the Courtyard in Hereford on the mezzanine floor.
If you attended our Software Freedom Day event, please feel to come along on Wednesday
The meeting will start at 7:30pm as usual.
Items to discuss :-
I had a great time at the weekend, taking photographs at Big Finish Day 4. The event is for fans of the audio production company, who make audio plays of Doctor Who, The Avengers, Dark Shadows, Blake’s 7 and much more. I’ve listened to Big Finish audio plays for years, mostly their Doctor Who range (of course!). The production standards are superb and one of their recent releases is up for a BBC audio drama award. I’ve been lucky enough to do some work for them over the last few months, and was asked to go along to capture some of the event.
In the morning I was wandering around taking candid shots of people enjoying the convention and the panels. It was rather like taking wedding photographs although slightly more relaxed. There are so many different moments to capture in a short time during a wedding ceremony, but a convention panel is a little more static and a good deal longer. Fortunately the urbane Nick Briggs kept the crowd laughing through the morning, and there was a really great atmosphere through the whole event.
In the afternoon I set up a portable studio to take some photos of various Big Finish actors. I was rather pleased with this set up, especially as it all managed it fit in my car! Apart from the background roll.
I really enjoyed working at Big Finish Day, catching up with some of the very nice people I’ve met at recording sessions, and hope to be asked back again!Pin It
Eat, Drink and talk LinuxEvent Date and Time: Wed, 22/01/2014 - 19:30 - 23:00
I have just run a search for further evidence of the possible compromise at thrustvps and found threads on webhostingtalk, vpsboard, freevps.us and habboxforum amongst others. All of those comments are from people (many, like me, ex-customers) who have received emails like the one I referred to below.
So, I guess thrust /do/ have a real problem.
In 2013 we kicked off Bad Voltage, a fun and irreverent podcast about technology, Open Source, gaming, politics, and anything else we find interesting. The show includes a veritable bounty of presenters including Stuart Langridge (LugRadio, Show Of Jaq), Bryan Lunduke (Linux Action Show), Jeremy Garcia (LinuxQuestions Podcast), and myself (LugRadio, Shot Of Jaq).
We have all podcasted quite a bit before and we know it takes a little while to really get into the groove, but things are really starting to gel in the show. We are all having a blast doing it, and it seems people are enjoying it.
If you haven’t given the show a whirl, I would love to encourage you to check out our most episode. In it we feature:
We also have a new community forum that is starting to get into its groove too. The forum is based on Discourse, so is a pleasure to use, and a really nice community is forming. We would love to welcome you too!
In 2014 we want to grow the show, refine our format, and grow our community around the world. Our goal here is that Bad Voltage becomes the perfect combination of informative but really fun to listen to. I have no doubt that our format and approach will continue to improve with each show. We also want to grow an awesome, inclusive, and diverse community of listeners too. Our goal is that people associate the Bad Voltage community as a fun, like-minded set of folks who chat together, play together, collaborate together, and enjoy the show together.
Here’s to a fun year with Bad Voltage and we hope you come and be a part of it.
I know it is futile to rant about banks. I know also that I should not really expect anything other than crap service from an industry that treats its customers as useful idiots. But yesterday I met with such appalling and unforgiveable stupidity and intransigence that I feel the need to rant here. I have left a period of 24 hours between the experience and the rant simply to allow myself time to reflect, shout at the cats and explain my frustration to my wife in the (vain) belief that it might actually be me at fault rather than the banks (I am, at heart, an eternal optimist).
Here is what happened.
Recently I was reviewing my meagre savings because the laughably small interest rate on one of my ISAs had been reduced following the ending of a “bonus period”. Great scam this. Offer a rate say 0.5% higher than that offered by competitors but limit it to a fixed period. Thereafter, drop the rate to (say) 1.0% below your competitors’ offerings. Better still, lock your customer into the agreement for a period (say) two years longer than the initial “high rate” period. We all fall for this, yet in my view it should be outlawed because it takes advantage of those least able to care for themselves – i.e people who do not actively manage their savings. Such people are often elderly, or infirm, and perhaps confused by the finance sector (who isn’t?). Some people are taken in simply because they naively believe that, as long term, loyal, customers they will be treated as such by their bank.
Following my latest review I noted that one of my ISA accounts (I have several, largely as a result of the aforementioned bank policy of varying introductory rates) was paying 3.0% when the one with the now lapsed introductory rate was paying only 2.0% (now to be reduced even further to 1.5% from 1 February this year). On the face of it would seem to be a no-brainer to move the funds in the lower paying account to the one paying the higher rate. But ISA investment is complicated by the need to make 2-3 year decisions now based on current rates when rates may rise later this year and by the fact that you are limited to a fixed maximum investment sum in any one financial year and in any one ISA. Decisions are further complicated by the policy of some banks to refuse transfers in from other (previous year) ISAs and the fact that some ISAs do not allow any withdrawals or indeed transfers out. Little wonder that “inertia” means that they get to keep your money at rates as low as a derisory 0.1% per annum. I kid you not. My wife was getting this from Barclays up until last year when I found out and stood over her whilst she completed the forms to move her cash. My wife is in the “loyal customer” camp.
So, do I move funds now in order to gain 1.0% or wait six months (when rates may have moved and we will in any case be in a new F/Y and sparkly new ISAs with fresh introductory carrots will be in place)? Tough call, but the fact the rate differential was set to rise to 1.5% in February, coupled with the fact that the Halifax (let’s name and shame) on-line banking system said that I could transfer in the ISA in question to their ISA paying 3.00% convinced me to start the process.
On starting the transfer I checked, and double checked, that the account I was transferring into was the one offering the 3.0% rate. It was. The account number and sort-code matched. On completion of the process, the system even gave me a nice “Thank you for transferring into your cash ISA” page to print which summarised the action I was authorising (“from” account detail, “to” account detail, amount of transfer, National Insurance No. etc). A couple of days later I noted that funds had disappeared from the “from” account so I knew the process must have started. I checked several times over the next week to see if the funds had been added to my Halifax account but was not overly worried that it had not because the Halifax does explain that the process can take up to 15 days, but that they will credit interest to the account for the full sum transferred from the date of application (the “Halifax ISA promise“). Well, I wasn’t worried until I received a letter from the Halifax saying that they could not complete the process of opening my account until I had provided information to verify my identity. Note that this letter confirmed the account details as being the existing ISA I hold with them. The letter concluded by saying that I should “call into any of our branches to confirm” the identity information requested. It also said that if I had any questions I should call a telephone number provided.
I have several accounts with the Halifax, including a mortgage account, but not a current account. My first mortgage in 1977 was with the Halifax. The Halifax has confirmed my identity often, and indeed fairly recently since the ISA in question was opened only one year ago. My nearest Halifax branch is a round trip of 25 miles away. I was at the time feeling pretty grumpy and in no mood to make an unnecessary 25 mile round trip to go through yet another identity check when I could call them to sort out the issue. (I should add as background here that I suffer from both arthritis and a form of inflammatory arthritis commonly known as gout. At the time I received the letter I was suffering from a gout attack. The normal treatment for pain relief in gout is NSAIDs such as aspirin, ibuprofen or diclofenac. I can’t take any of those pain killers because I am allergic to the damned things. The main drug prescribed for relief of gout inflammation is called colchicine. Colchicine has an interesting set of side effects (look them up). Now figure out why I was grumpy and disinclined to make an unnecessary 25 mile round trip.)
I called the 0845 number only to be greeted by the monumental stupidity of one of those automated systems which requires entry of a specific type of number before it will take you any further. Such systems seem to have been designed specifically to wind you up to the stage that you give up and go away. Mistakes numbers 1 and 2 here are 1, refer customers to such a number on a letter which says “call here if you wish to discuss any questions arising from this letter” and 2, assume that all callers to this number will have current accounts and thus possess the magic number required to use the system successfully. Eventually, however, through persistence, I got to a human being. The human being in question appeared to be a nice helpful young man in a call centre. Unfortunately after listening to my explanation as to why I was calling, the NHYM in question said he was “not trained to answer” my particular question. I know that such calls are recorded, but I found the precise wording he used rather odd, obviously part of a script he was forced to follow in certain cases. I guess that there are certain key phrases which must be used in all transactions for regulatory reasons. As an aside, I find it sad that even in the situation where you do eventually get to speak to a human being after going through a brain dead automated process, that human being is then forced to act and speak like an automaton.
NHYM number one then passed me back to the automated system whilst reassuring me that he was actually passing me to a colleague who would be able to respond to my particular problem. Whilst I was on hold, the automated system asked me if I was aware that I could get nearly all my questions answered by going on-line to the Halifax web site. Mistake number 3 lies in assuming that anyone who prefers to talk to a human being on a telephone is remotely interested in (or maybe even capable of) using a web service. Those people who have got to the telephone are either web users who have found that the site does /not/ meet their requirements or they are people who actively prefer not to use a website. Referring either group to the web is thus both counter productive and irritating.
Having been on hold for a short period and having ignored all requests to press button “X” or “Y” I was again eventually connected to a (different) NHYM in a call centre. I then again explained my situation: viz. I had received a letter asking me to schlepp up to my nearest Halifax in order to provide documentary proof that I was the same bloke who had opened a particular ISA one year ago. I pointed out that the Halifax knew who I was from previous encounters (I listed various accounts in evidence) and that it should be relatively easy to confirm this on the telephone – hence my call. NHYM number two then took me through various “security questions” which will be familiar to anyone who has a UK bank account (give full name, confirm first line of address and post code, give balance of account in question, give National Insurance Number (it is an ISA), state date of birth). Having done all this, NHYM number two then thanked me and said that he would arrange for me to receive an identifying number which I could use in future telephone banking interactions which would prevent me having to go through this rigmarole all over again. However, he said, I would still have to go into my nearest Halifax branch to prove my identity because this is a “legal requirement” and part of “our know your customer programme”.
I confess that by this time my patience was stretched a little thin and I may have been somewhat abrupt with the NHYM in expressing both my incredulity that this should still be necessary and my intense irritation that the Halifax should assume it OK to insist on my travelling to them when I have already been through the necessary hoops more than once. I pointed out that I had just gone through the process of proving I was who I said I was to his satisfaction in order to meet their security requirements. I also pointed out that for all he knew, I could be a disabled old man incapable of leaving his home without assistance (not actually true, but not so far from the truth) and that it was therefore a little unreasonable to expect me to do so. NHYM sympathised, said that he would pass on my complaint, but felt obliged to point out that any formal complaint would not be upheld because Halifax was merely obeying its legal obligations to identify its customers and the source of their funds. (I know for a fact that this is utter bollocks and that banks choose how they should meet their obligations under money laundering regulations, but I did not feel that this would be a fruitful line to take with the NHYM so I limited myself to forcefully asking him to lodge my complaint). Before concluding our conversation, I apologised to the NHYM for venting my frustration upon him and told him that I realised that he was following instructions in a difficult job and that he was in no way to feel himself at fault for the failings of his employer and its systems.
I then made the trip to my nearest Halifax. Before going however, I decided that I should take the opportunity of the visit to close an old branch book based savings account I hold as trustee for one of my grandsons and transfer the balance to a newer account I hold elsewhere which offers a much better rate (they play the same “bonus interest rate” trick on poor defenceless children.)
On arrival I was greeted by a NHYL who listened to my story and then told me that what I had described could not possibly have happened because I was not allowed to transfer in new funds to an existing fixed rate ISA account. What I must have done (according to her) was to open a new ISA and request transfer of funds into that. I showed her the documentary evidence refuting this statement and also pointed out that the money had obviously been requested by the Halifax because it had left my other account about a week ago, NHYL again said that this was not possible because the Halifax would not request the funds until the account had been set up. The fact that the letter I held in my hand said that the account could not be opened until they had verified my identity proved that the account had not yet been set up. I pointed to the account number and pointed out that this number was my existing ISA account which was already open, so perhaps the Halifax had requested the funds in order to place them in my existing account. Again, NHYL said this was impossible. I then asked where my money was because it clearly was not in my other account and the Halifax appeared not to have it. She proved unable to answer this and suggested that I check with my other bankers as to what they had done with the money. So I trotted (slowly) up the road to the other bank and spoke to yet another NHYL (though this one actually /was/ helpful). Apparently, despite the Halifax NHYL’s protestations to the contrary, the Halifax had requested transfer of my ISA funds almost immediately. My bankers had simply responded as they were expected to do and passed the money on. On hearing my full story, and learning that I did not now wish to pass my funds to the Halifax (old account or not) NHYL number two spoke to her colleagues in the bank’s ISA section and confirmed that once the funds were returned to them (as they would have to be if the account opening process was not completed) my old ISA would be re-instated.
I thanked the helpful lady and made my way back to the Halifax where I updated the Halifax version of the NHYL on the position. Whilst there I also asked how I should go about closing my grandson’s account. I was directed to the counter. At the counter I entered what appeared to be yet another banking bubble divorced from reality. My request to close the account and let me have a cheque made out in my grandson’s name was greeted with the response that it would not be possible to provide a cheque for a sum less than £500.00. I was offered cash. I said that I would prefer not to take cash because (bizarre as it may sound) I knew that the bank holding my grandson’s account did not accept cash, dealing only in cheques. (Sometimes I despair). I then asked if the cashier could just transfer the balance of the account to my grandson’s other account and then close the old account. Reader, you will not be surprised to learn that this proved impossible.
I took the cash and left.
As I said at the start of this rant, I really should know better. Despite all the (supposed) huge investment in technology in our banking systems, systems which we rely upon in ways which are deeply fundamental to our society, those systems consistently fail to meet quite simple needs. Frankly, this terrifies me.
Note carefully that all of the problems relating to my ISA transfer described above could have been prevented very simply. All that was necessary was a check in the on-line system which should have popped up a warning when I attempted to transfer funds saying “Sorry, you cannot transfer funds into this account. Do you wish to open a new ISA?”
To which my answer would have been: “No”.
Which is better … streaming content or buying shiny plastic discs?
It’s surely an unfair comparison because the constraints of a service like Netflix (delivering uninterrupted video over a range of network qualities) are very different to delivering content on physical media. But I was looking to justify my purchase of Battlestar Galactica (BSG) on Blu-Ray so here’s some screen captures as a very unscientific comparison. The screen captures are only approximately the same frame on each media, but should be close enough for a rough comparison.
The Netflix screen captures were done streaming on a fibre internet connection using a Mac with Silverlight installed, but with no other adjustments. The Battlestar Galactica Blu-Ray claims to be “High Definition Widescreen 1.78:1” (but doesn’t define high definition further … is it supposed to be 720p or 1080p?).
Here’s the two sources at 100% magnification (click for full-size):
And at 200% magnification:
And at 300% magnification:
In this instance, Netflix does a pretty good job, although it looks a little blurry. BSG is probably a bad choice for showing off Blu-Ray in comparisons: it’s heavily processed to make it look grainy like film, and almost every single scene is very dark. I doubt it was shot on the latest HD cameras, either (the miniseries first aired in 2003, the full series began in 2004). But since more recent series like Game of Thrones aren’t available on Netflix, this is the best I’ve got for comparisons at the moment.
Of course, even if the Blu-Ray and Netflix versions were identical, a big benefit of those shiny plastic discs is they can be used off-line and don’t require a continuing subscription.
Subject: Damn::VPS aka Thrust::VPS
Date: Sat, 18 Jan 2014 03:28:06 +0000
This is a notification to let you know that we need to verify for reduce fraud.
We want your data as soon as possible.
The data that we need is as follows:
Server Username (Included)
Server Password (Included)
Full Name (Included)
Phone Number (For Call To Verify)
Paypal Email (If Order With Paypal)
Paypal Password (If Order With Paypal)
Credit Card Information (If Order With Credit Card)
Scan Of Credit Card Front And Back (If Order With Credit Card)
Data is sent to Email : email@example.com
Thanks in advance for your patience and support.
http://damnvps.com – Damn::VPS – We give a damn
Now, apart from the fairly obvious phishing nature of this email (you want me to scan my credit card front and back and send you a picture? Right…), and the request to reply to an address other than the sender (“Data is sent to….”) it actually looks to me as if the email really came from Thrust. Certainly the full headers (including “Return-Path:”, “Reply-To:”, “Received:” and even “Message-Id:”) look remarkably similar to the real ones I have from earlier mails from Thrust. A normal phishing email will usually spoof the “From:” address and use the “Reply-To:” to capture return emails at the scammer’s address. The fact that this email actually asks (in grammatically poor english) that you reply to a yahoo address suggests that the scammers are not that sophisticated.
I may not have much time for Thrust, but I have even less time for spammers and scammers so I forwarded the email to Thrust with a recommendation that they check it out and let their customers know that there appeared to be a scam going on in their name. I also checked their website to see if they had any alert thereon. The website (as at 15.00 today) appears to be unreachable (and I have tested from the UK, SanFrancisco, NYC and Amsterdam). With a website down and dodgy mail appearing to come from a legitimate Thrust mailserver address it suggests to me that they may have suffered a compromise. Certainly it looks to me as if their customer email database has been compromised (the address I got the email on was not my normal address, rather it was the one I use for contacts such as this). Whether that means any of their other account details have also been stolen I cannot be sure.
But I am glad that I am no longer a customer.
Now it is time for me to go silent for a while, and not talk about jobs, unemployment, or puppies.
This past week has also been full of software releases. Some of the public ones include:
After three months of slow work I've issued a new release today. This release features several bugfixes for dealing with malformed MIME messages, and similar fun.
The core set of lua primitives hasn't changed very much for a good six months now, which means I guess rightly what kind of things would be useful.
This was recently updated to add two new plugins to the core:
Although there are a million static-site generators I still think mine has value, and I am consistently using it.
Months ago when I said "I'm writing a mail-client", all I need to do is handle three cases:
Then some new things like "Compose", "Reply", "Forward", I remember somebody commented along the lines of "Yeah, but MIME will make you hate your life" I laughed. Now I know better. Still it works, it works well, and I'm glad I did it.
I was back at my parents' over Christmas, like usual. Before I got back my Dad had mentioned they'd been having ADSL stability issues. Previously I'd noticed some issues with keeping a connection up for more than a couple of days, but it had got so bad he was noticing problems during the day. The eventual resolution isn't going to surprise anyone who's dealt with these things before, but I went through a number of steps to try and improve things.
Firstly, I arranged for a new router to be delivered before I got back. My old Netgear DG834G was still in use and while it didn't seem to have been the problem I'd been meaning to get something with 802.11n instead of the 802.11g it supports for a while. I ended up with a TP-Link TD-W8980, which has dual band wifi, ADSL2+, GigE switch and looked to have some basic OpenWRT support in case I want to play with that in the future. Switching over was relatively simple and as part of that procedure I also switched the ADSL microfilter in use (I've seen these fail before with no apparent cause).
Once the new router was up I looked at trying to get some line statistics from it. Unfortunately although it supports SNMP I found it didn't provide the ADSL MIB, meaning I ended up doing some web scraping to get the upstream/downstream sync rates/SNR/attenuation details. Examination of these over the first day indicated an excessive amount of noise on the line. The ISP offer the ability in their web interface to change the target SNR for the line. I increased this from 6db to 9db in the hope of some extra stability. This resulted in a 2Mb/s drop in the sync speed for the line, but as this brought it down to 18Mb/s I wasn't too worried about that.
Watching the stats for a further few days indicated that there were still regular periods of excessive noise, so I removed the faceplate from the NTE5 master BT socket, removing all extensions from the line. This resulted in regaining the 2Mb/s that had been lost from increasing the SNR target, and after watching the line for a few days confirmed that it had significantly decreased the noise levels. It turned out that the old external ringer that was already present on the line when my parents' moved in was still connected, although it had stopped working some time previously. Also there was an unused and much spliced extension in place. Removed both of these and replacing the NTE5 faceplate led to a line that was still stable. At the time of writing the connection has been up since before the new year, significantly longer than it had managed for some time.
As I said at the start I doubt this comes as a surprise to anyone who's dealt with this sort of line issue before. It wasn't particularly surprising to me (other than the level of the noise present), but I went through each of the steps to try and be sure that I had isolated the root cause and could be sure things were actually better. It turned out that doing the screen scraping and graphing the results was a good way to verify this. Observe:
The blue/red lines indicate the SNR for the upstream and downstream links - the initial lower area is when this was set to a 6db target, then later is a 9db target. Green are the forward error correction errors divided by 100 (to make everything fit better on the same graph). These are correctable, but still indicate issues. Yellow are CRC errors, indicating something that actually caused a problem. They can be clearly seen to correlate with the FEC errors, which makes sense. Notice the huge difference removing the extensions makes to both of these numbers. Also notice just how clear graphing the data makes things - it was easy to show my parents' the graph and indicate how things had been improved and should thus be better.