503 Service Unavailable in US East July 3?


  • Is anyone else getting 503 Service Unavailable errors for Azure Web Sites today, July 3? We've been getting them intermittently all day across all of our hosted sites, but there's nothing posted on the Azure Status dashboard. In the past I've been able to rectify this by scaling up the number of servers for our Standard Plan and then scaling back down, which seems to force a reset. That seemed to work this morning but isn't working now.
    Thursday, July 03, 2014 6:45 PM

All replies

  • Here I getting 503 too

    MCP, MCTS 2.0 Web Applications, MCTS 3.5 ASP.NET

    Thursday, July 03, 2014 7:04 PM
  • We don't have paid support, which seems to prevent us from even reporting this issue. Any idea if there's a way to report a possible outage without paying for support?
    Thursday, July 03, 2014 7:14 PM
  • Web sites team is already working on this issue and making sure it won't happen again.

    There was information posted on

    In general, we monitor this forum pretty closely as well, so even though there is no guaranteed support, we look into issues reported here pretty quickly as well, so thank you for posting it (but we have our internal monitoring reporting the issue too).

    Apologies for the inconvenience.


    Thursday, July 03, 2014 8:19 PM
  • Hi Petr,

    The problem still here!

    MCP, MCTS 2.0 Web Applications, MCTS 3.5 ASP.NET

    Thursday, July 03, 2014 9:11 PM
  • Hi Petr -

    Thanks for the update. The information on the status portal was posted many hours after the issue began for us (8:58 AM today). We had additional outages this afternoon around 1 PM.

    It's good to know for the future that you're monitoring the forum for such issues, since there doesn't appear to be a way to report potential system outages.

    Thursday, July 03, 2014 9:40 PM
  • We just had an intermittent 503 on a couple of our sites in US East again a few minutes ago. They appear to be back up now, but there's nothing on the Status Board.
    Monday, July 07, 2014 4:38 PM
  • We don't see any wide issues at this moment. Would you be able to provide us with a site name so we could look further? There is also a possibility that this could be an application level issue.


    Monday, July 07, 2014 5:18 PM
  • It was multiple sites simultaneously that are completely different applications, so I don't think it's an application issue. It's also happening on some deployments that haven't been updated anytime recently. If you can provide an email address I'll send you a list of some of the affected sites.
    Monday, July 07, 2014 5:20 PM
  • Can you please send that list of sites to "websites at microsoft dot com"? Please include timestamp of the issue (from your post it looks like it was around 4:30 PM UTC time, but I would like to confirm).


    Monday, July 07, 2014 5:44 PM
  • Sent!
    Monday, July 07, 2014 5:57 PM
  • This was truly an application issue. Sites used all available memory (at its peak there was 3.3 GBs committed and a small VM in Azure has only 1.75 GB memory) on the assigned machine which became really slow. Scaling to multiple machines does not solve this well. You need to scale to larger machines (instead of two smalls, one medium should help e.g.).

    You should be able also to observe the memory usage in your Azure management portal.


    Monday, July 07, 2014 8:55 PM
  • Hi Petr -

    I see where the memory usage per site is, but is there somewhere in the Portal that lists it across all sites on the VM?

    Monday, July 07, 2014 9:41 PM
  • You can see three colors there. Green represents your site, dark gray all your other sites in the same hosting plan (sharing the same set of VMs) and light gray is for available.

    If you are asking for a breakdown per site in form of { SiteA 123 MBs, SiteB 234 MBs and SiteC 456 MBs }, then I'm afraid it is not there today. It should be possible to retrieve such data through our API though.

    Hope that helps.


    Monday, July 07, 2014 10:11 PM
  • Hi Petr - I'm not seeing where there's a graph like that. In the old Portal navigating to a site's Monitor tab has usage graphs, but there doesn't seem to be anything matching the colors or data you're describing, even with the memory-related Metrics enabled. In the new Portal I do see there's a memory quota usage for the Server Farm, which does seem to have the current day's usage across sites, although it seems to lack the option to switch to a 1-hour or 7-day view like in the old Portal.

    Can you tell me how to get to the graph you're describing? Thanks!

    Tuesday, July 08, 2014 9:08 PM