Pingdom Home

US + international: +1-212-796-6890

SE + international: +46-21-480-0920

Business hours 3 am-11:30 am EST (Mon-Fri).

Do you know if your website is up right now? We do! LEARN MORE

Twitter growing pains cause lots of downtime in 2007

Twitter logoIn most ways, 2007 has been a great year for Twitter. The service has grown tremendously and has become one of the big social network successes.

The flip side of this popularity is that they have often been unable to handle all the traffic their large user base has been bringing in, which has affected the stability of their website. In other words: Twitter has been suffering from growing pains.

We have monitored the availability of the Twitter.com website since mid February 2007 until today (and counting). So far, the accumulated downtime of the website totals 5 days and 23 hours. This includes maintenance, but is still a very high number. So far in December, the website has been unavailable for a total of almost 11 hours (part of which was due to problems related to a data center move).

The worst month for Twitter by far was March, when the Twitter.com website totaled over 2 days and 10 hours of downtime. This is also the month when Twitter’s popularity really took off, winning the SXSW Interactive Festival, so it looks like the rapid growth that month took its toll.

Twitter.com downtime in 2007
Month Uptime (%) Downtime (d, h, m)
December 2007 97.51% 10 hours, 59 mins
November 2007 98.74% 9 hours, 4 mins
October 2007 99.02% 7 hours, 19 mins
September 2007 99.06% 6 hours, 44 mins
August 2007 98.47% 11 hours, 22 mins
July 2007 99.59% 3 hours, 2 mins
June 2007 99.08% 6 hours, 37 mins
May 2007 97.67% 17 hours, 9 mins
April 2007 99.11% 6 hours, 15 mins
March 2007 92.09% 2 day, 10 hours, 3 mins
February 2007 98.21% 6 hours, 22 mins
TOTAL 98.06% 5 days, 23 hours, 0 mins

Above: Note that the February data is not complete, and that the December data of course isn’t complete as of this writing. This data can be found, always up to date, on the public uptime report page for Twitter.com. (A day-by-day breakdown can be had by clicking on the link for a month.)

Another indication that the Twitter.com website has been suffering from more traffic than it can handle is that a lot of days the website has been unavailable for numerous short periods of time, often ranging from 1 to 10 minutes, spread over large parts of the day, which indicates that the website has been struggling to keep up with demand and deliver pages during much of that time.

This last weekend Twitter moved their service to a new data center, which will hopefully alleviate the problems they have been experiencing.

Will the Twitter website be more reliable from now on? As we mentioned, you can keep track of this on a public report page from Pingdom that shows the uptime of the Twitter.com website and is updated continuously.

Note: Twitter’s recent downtime and data center move has received a lot of buzz and speculation, for example in Mashable, Ensight and Data Center Knowledge.



0 comments

Trackbacks

  1. [...] them – service was down for a total of 2 days, 10 hours and three minutes. (Link) Share/Send Sphere Print PreviousPost [...]

  2. [...] Twitter, the micro-blogging service that gained momentum with the SXSW Interactive Festival back in March, exploded in popularity this year. And with that came a sharp increase in the outages. Pingdom says Twitter was down for about six days in 2007. No surprise that March was the worst month for them — the service was down for a total of two days, 10 hours and three minutes. (Link) [...]

  3. [...] Even after moving to a new data center, Twitter has been undergoing some growing pains. Royal Pingdom reports that the service has been down almost a total of six days since it launched last February. Downtime so far in December has exceeded almost any preceding month, with nearly 11 hours of downtime, compared to 9 hours in November. [...]

  4. [...] Even after moving to a new data center, Twitter has been undergoing some growing pains. Royal Pingdom reports that the service has been down almost a total of six days since it began monitoring the service last February. Downtime so far in December has exceeded almost any preceding month, with nearly 11 hours of downtime, compared to 9 hours in November. [...]

  5. [...] Posted: 20 Dec 2007 08:49 AM CST Even after moving to a new data center, Twitter has been undergoing some growing pains. Royal Pingdom reports that the service has been down almost a total of six days since it launched last February. Downtime so far in December has exceeded almost any preceding month, with nearly 11 hours of downtime, compared to 9 hours in November. [...]

  6. [...] The list includes Skype Outage, Twitter growing pains, ubuntu.com crash, Google Analytics break. Check out pingdom to see the notable internet outages in 2007. [...]

  7. [...] According to Royal Pingdom, Twitter was down for almost 6 days (5 days and 23 hours to be exact) in 2007. For any other company, this might spell doom, but not our LittleCo of the year! If anything, the service has gotten more popular in spite of the frequent down time. Those with “Twitterdiction” may have suffered through some nasty withdrawal when Twitter went offline for extended periods of time, but abandon the service they did not. [...]

  8. [...] In late December 2007 – the blogosphere lit afire with the story of downtime statistics for the hot microblogging service Twitter. This was reported by pingdom on December 19th. For those not familiar with microblogging, Jermiah Owyang provides a great primer on Twitter. (If you Twitter, you can follow me here.) [...]

  9. [...] Downtime has become par for the course for Twitter during big news events (especially those pertaining to the tech industry, which makes up a sizable portion of the Twitter userbase), with the company having nearly 6 days of downtime in 2007 according to a recent report from Pingdom. [...]

  10. [...] Ces “downtime” sont devenus assez courants sur Twitter surtout lors d’évènements et conférences (spécialement celles qui intéressent la cible Tech, qui représente une très large part des utilisateurs), d’après un récent rapport de Pingdom, le site aurait même enregistré +6 jours d’indisponibilité en 2007. Pourtant, le service prend vraiment tout son sens lors de tels événements qui représentent un réel intérêt pour ses membres. [...]

  11. [...] Но закрывая тему яблок хочется рассказать о том, что в момент презентации продукции Apple, они называют ее Keynote, каждый уважающий себя пользователь компьютеров данной марки и пользователь Twitter по совместительству выдавал в Twitter такое количество эмоций, что тот не выдержал и начал то падать то вставать, что для него обычное дело — каждый подобный Hi-Tech event данный сервис является лакмусом интереса. Если интересно, то он просто падает :) Забавно, что на другие не-IT новости он реагирует стойки и надёжно. Кстати, по поводу стабильности Твиттера были проведены исследования и получилось, что бедняга в 2007 году в сумме лежал и дрыгал лапками 5 дней и 23 часа — почти неделю :) [...]

  12. [...] Entire month-long online conversations occur when an uptime monitoring service posts data showing that one quickly growing popular web 2.0 service had over 6 days of downtime in 2007. [...]

  13. [...] Unos días atrás, escribía sobre Twitter y el impacto que ha tenido la aplicación a nivel mundial durante el último año. Aunque no lo comenté en aquel entonces, mi motivación principal para comenzar a “tuitiar” era tratar de integrar un sistema como Twitter, flexible, para nuestra oficina. Sin embargo, las limitantes propias de la aplicación aún no lo permitían (por ejemplo, no permite mensajes privados a grupos de usuarios específicos, los contenidos no pueden ser categorizados mediante tags o por canales temáticos, dada la demanda el servidor de Twitter le ha costado mantenerse en un 100% en línea, entre otros) [...]

  14. [...] La grande majorité des utilisateurs de Twitter aiment ce service addictif au détail près qu’il est d’une instabilité exemplaire. Pingdom avait déjà noté le nombre de ruptures de service sur l’an 2007 et on aurait pu croire qu’avec les fonds d’investisseurs et leur nouveau CTO en place ces problèmes auraient disparu. Cela a continué lors de la MacWorld Expo ou le service n’a pas tenu et sur les dernières 24h00 c’est la débâcle; le service est totalement inaccessible. Les utilisateurs sont révoltés (moi le premier), Evan Williams est obligé de s’excuser sur le blog de twitter. [...]

  15. [...] The outer limits of scaling – how decentralized can your infrastructure be? March 11, 2008 at 8:45 pm | In mashups | Tags: infrastructure, on-demand, ondemand, utility computing, utilitycomputing I’m a big Twitter fan. So it’s painful to see that Twitter is down so often. This problem with scaling from niche service to Internet starlet has plagued a good number of other young (and not so young) websites. The scaling problem itself (growing without experiencing significant performance issues) goes right back to the earliest Internet startups – if you can’t keep adding servers to meet the demand of the people clamouring to get on your site, you’re screwed. [...]

  16. [...] Twitter seems to have better uptime than most of the financial institutions. Guess I’ll be rebalancing another night… Post a comment | Trackback URI [...]

  17. [...] il 2007, Twitter è stato irragiungibile per 5 giorni e 23 ore. 143 ore di "buco", durante i quali gli utenti non hanno visto altro che [...]

  18. [...] A site that monitored Twitter uptime in [...]

  19. [...] his getting tiresome. See Dave’s Twitter uptime monitor of May 23, 2008 vs. Pingdom’s report from Dec 19, 2007. Also see Dave’s decentralizing Twitter “idea” from May 4, 2008 [...]

  20. [...] load.  These periods of “downtime” last for anything from 5 minutes to 2 days (and 5 days in 2007). You’d have thought they’d have solved the issues by now – especially considering [...]

  21. [...] The Bezos addition is interesting because if anyone knows how to scale an online business it’s Amazon — and Twitter definitely needs help scaling because despite (or probably more likely because of) its popularity the service has suffered repeated downtime. [...]

  22. [...] the inaugural launch year of  Twitter, the company had over 5 days of downtime in the entire year according to Pingdom.com, now when a community is relying on this site for “real time conversations” 5 days is a [...]

  23. [...] la gara dei social network più inaffidabili. In questo sito, il numero di ore in cui è andato in downtime nel 2007 suddiviso per mesi . In quest’altro link, il downtime da febbraio 2007 ad oggi. C’è [...]