1. Pingback: Paul Marshall
  2. Pingback: Rachel Reeves
  3. Pingback: Frances Hoggarth
  4. Pingback: Alan Marshall
  5. Pingback: davidschoibl
  6. Pingback: Lawrence Mills
  7. Pingback: Samuel Tarry
  8. Pingback: IKP
  9. Pingback: Sue Thomas
  10. Pingback: Tracey B
  11. Pingback: Chris DeRose
  12. Pingback: pueblo harrass
  13. Pingback: Emily Dubourg
  14. Pingback: Emily Dubourg
  15. Pingback: Keef Hinton
  16. Pingback: Liz W
  17. Pingback: Colin Harris
  18. Pingback: star_tourmaline
  19. Pingback: David Derbyshire
  20. Come on- this is obscene political warfare- “our site has crashed due to new members”…. how convenient and how brilliant as a press release.

    Proof? A screenshot…

    Perhaps you should be a proper journalist and ask labour for their official analytic data and have it independently checked?

    Oh no, you’re a blogger so you wont bother.

  21. Pingback: scott somerville
  22. I have to agree with Dixon on this one; it has a definite smell of ‘Campbell-ness’ about the announcement. I can only guess that the real proof will be borne out at the next General Election.

    Anyhow, I’ll go back to drumming my fingers on the desk until I see undoctored, live analytics from the Noo Labour website.

  23. The likelihood of the website programmers having put this page in as an error handler is very close to nil – any reasonable website should be able to handle 10’s of submissions per second, let alone one every 3 seconds (20 people divided by 60 seconds). Looks like a static page put up for spin purposes to me! (I’m a web programmer)

  24. Yup, my really low spec web server can handle hundreds of requests *at the same time*, let alone per minute. “A record number of people” is obviously very choice words.

  25. Just goes to show, that even in defeat, they can’t just crawl off with their tails between their legs, they have to ‘spin’ their way back under that rock!! Well done Richard for showing us the real site at that time.
    No web admin would put a error page up like that, there are many reasons you could get an error, and if this message was really therem they could just drop the connection to the backend to reproduce it at will.

  26. I was trying to join all last night and couldn’t because of traffic. I didn’t get the screen shot posted above, but I did get a message about the server being too busy and to try again later.

  27. I’m a web dev. Technically I mostly develop OO Applications which integrate web and desktop, but I’d like to confirm that when I joined up 2 days ago the server was already straining under the load.

    My application went through ok, but at the final confirmation point the dialog box hung. Being a techie I started wondering why – and whether it had actually been processed. I poked about with my various developer debug tools (the developer console in Safari is very useful) and the server had indeed failed to respond to one final script due to weight of traffic. Well – it was an internal server error, but 9 times out of 10 that’s ‘whoa, I just don’t have the capacity to run this right now’.

    Obviously I can’t go back in time and capture a screen shot for the terminally cynical. But if it happens again when my partner joins up I’ll be sure to do so.

    The original error message was much less attractive. I imagine that their server had a maximum threads / ram setting (possibly in the php config) to alert them to overload through a denial of service attack. The attractive error message above is a temporary manual page slammed up while the server config was adjusted to increase capacity to cope. It’s not a conspiracy, it’s normal practice in my business. We had to do the exact same thing recently when a client’s product was featured on Oprah. We don’t routinely give a ton of resources to apps / processes that normal only need a teaspoon.

    You do sound ridiculous claiming this is spin! I imagine both were true – some people got the hideous 500 type error and some got the nicer looking ‘there’s a problem’ page (which will have been manually created this morning – it looks like, what, 90 seconds of work?) That’s what tends to happen when your server is being adjusted.

  28. Let me just say, I’m a web programmer.

    Even the cheapest £2.99 web hosting package would allow for at least this number of visits / database transactions per second.

    This is pure spin, and shame on BBC News for regurgitating this as ‘fact’

  29. Pingback: Jonty Hall
  30. Pingback: bigJohn Noble
  31. Jamie, it would allow it, but you’ve no idea how many other threads the site is also running, whether the hosting package is split with other applications, or whether they’ve put a modest limit on manually to alert them to DOS attacks. The error was real – and it was traffic that caused it.

  32. “it was an internal server error, but 9 times out of 10 that’s ‘whoa, I just don’t have the capacity to run this right now’.”

    No, it isn’t. Not even close. That error can be generated by a million and one things. An incorrectly configured distro, an error in their ASP code if using MS products, a database server that hasn’t had its log file cleaned out (if using SQL Server, that is) and that’s just three. On Unix, add mis-set permissions and incorrectly capitalized filenames to the mix.

    And a fake screenshot is a bit strange to us as fact. Shame on the BBC.

    Source: Been a web developer since 1995.

  33. Oh, and given that they’re using HTTPS for signups, it could even be that their HTTPS certificate server is incorrectly configured to only issue a certain amount of certificates at one time. The default is 24, and if they aren’t being destroyed cleanly, that’ll get used up very fast indeed.

    Hell, could even have been a DDOS attack as you mentioned. Doesn’t mean that people are ACTUALLY signing up!

  34. Stray does put forward a convincing argument. But I would like to know where the quote “Twenty people per minute” came from. Since that small amount of people should barely affect the server, even if there is much traffic elsewhere on the site.

    Perhaps it is the blogger’s text that is fiction, not the Labour party’s statement. And the load may be caused by general visits to the labour party website and not necessarily by people trying to join the party?

  35. I got the error when attempting to join last night; it did go through finally, but the website was slow and unresponsive – heavy traffic? What’s the capacity of the site? Who knows.

    I was also on an election thread on a football fans forum last night. Within an hour or two of the completion of the Con-Dem coalition, 6 or 7 people on that one forum confirmed they had joined (or re-joined) the Labour Party. 6 or 7 does not a deluge make; but from a single, unrelated, football fans forum? There does seem to be some genuine momentum to this.

  36. Pingback: Mark Kerrison
  37. Pingback: Edwin Ford
  38. At least a quarter of my Facebook friends have joined the Labour Party in the last two days. I suspect many more have too.

  39. To all the naysayers: yawn. The party has no money. That the technical infrastructure would strain under the weight of a hit-spike is entirely plausible.

    While it does make for a good little story for the party it isn’t spin. All indications are that, in addition to displaying errors, the server was sluggish/unresponsive. Are you suggesting that this degradation in performance was simulated by Labour’s technical wizards with Bad Al Campbell standing over their shoulder?

    As for “Yeah but I’ve been a web developer for X years and …” Ahem. Please see Stray’s comments above.

  40. Whether the server gave the message for the “spinned” reason or another, I’m more intrigued why so many, at least on the evidence of this comment section, are flocking to join Labour at this stage. WHy not lasy week, last month, etc? Apart from being in opposition & Broon having left, what has made your mind up?

  41. Maybe Brown’s departure was the reason – you never miss what u have till it’s gone, but maybe people who consider themselves progressive (i.e. NOT Tory) and who were duped by the Liberal Democrats simply wanted to realign their affiliations?

    Admittedly on the morning of 07 Jun 2010, I had no difficulties signing up when I returned home, but then that was before Nick Clegg had actually betrayed the strategic interests of those progressives who had loaned him their votes.

  42. @Political Scrapbook

    A little late, I know, but I’ve been doing a little digging on the matter for my own blog.

    Seeing as the traffic statistics point to May 11th being a very normal day for the Labour website in terms of traffic (with there actually being a much bigger spike earlier in the month with no problems reported), will you now admit that this incident was just lies and spin of the highest order?

  43. @James Laughlin

    Nice try but it could be a little bit more complicated than you realise. The process of signing up new members will rely on several systems working together in a coordinated way. As Ken points out above:

    “Oh, and given that they’re using HTTPS for signups, it could even be that their HTTPS certificate server is incorrectly configured to only issue a certain amount of certificates at one time. The default is 24, and if they aren’t being destroyed cleanly, that’ll get used up very fast indeed.”

    None if this changes the fact that Labour signed up 10,000 people in a week.

Leave a comment

Your email address will not be published.


Comments are limited to 1000 characters.