The provider we use for SMS is making changes on how Sprint traffic is handled. This is resulting in all Sprint users getting a error 2115 when trying to send to 40404. 

We will update here as we get more information as to when the problem will be fixed.

Update (12:55p): Our provider has resolved the connectivity issue in the past hour or so, but we are still working with users who reported the problem to make sure it’s been fixed.

We introduced a bug into the facebook app which we are working to fix.

Update (6p): This has been resolved.

The network issues that caused the search outage have been corrected. Search is again available with dat during the outage missing.

There’s a network outage affecting the data center where search.twitter.com is hosted. We are working with our hosting provider to resolve this as quickly as possible.

search.twitter.com is about ten minutes behind realtime.  This should be corrected shortly.

Update: this has been corrected

The outages we experienced today and last night were caused by a sub-optimal configuration of our database servers (a problem that we knew about but did not expect to go critical before it was resolved).

We’re making changes based on the outage to improve performance. New machines will go online as early as this evening. We do not anticipate the need for planned maintenance in order to make these changes. Should that change, we will update on the site and here as well.

User restoration and deletion (computationally intensive functions) remain offline as we work through these issues.

We’re recovering from an outage and will post here with updates.

Update (9:49a): We’re back up and will be conducting an extensive review of the problem that happened this morning and the one that occurred last night. We apologize for the unplanned downtime.

We uncovered and fixed a major problem with the SMS service that resulted in a large number of messages being dropped. This problem affected users from all carriers. 

There may still be problems that remain. We know there’s additional things that we need to do to make the system more reliable. However, fixing this problems should result in a significant improvement in message deliverability.

Status deletion has been re-enabled.

User restoration has also been re-enabled.

User deletion remains off as it still represents a significant problem for the stability of the service. In the past several months, we’ve had to turn off user deletion many times in order to keep things running smoothly. Each time we turned it back on, we’d hoped we had sufficiently patched the existing system to prevent it from being a problem.

We’ve now concluded that the user deletion system needs to be rewritten and we’ve started that effort. Until we finish, user deletion will remain off. There’s a number of places throughout the site where we will need to message this better. We’ll be making those changes as well.

There are a couple errors happening with image upload for profile pictures. Some images become resized incorrectly and break formatting when uploaded. Others are getting errors when uploading images (particularly with gifs) even though the files conform to our size restrictions (under 700k).

We are reworking the image upload system to deal with both of these problems (and others).

We’re experiencing a site outage and are working to recover.

Update (1:32p): And we’re back - this was a DNS configuration error that impacted the entire site.

We apologize for the unscheduled downtime. We’ve not had an outage of this length since mid-July and will be carefully reviewing what went wrong.

The search.twitter.com feed is behind.  This will be resolved shortly.

Update: this has been resolved

We’ve temporarily disabled user deletion and restoration to speed the recovery of the data inconsistency issues noted previously. 

We had a database replication service fail last night, which has caused a number of problems — missing direct messages, direct message counts off, missing replies, etc. All the data is in the system, it’s just not in the right place.

It’s now in the course of correcting itself but will take an hour or two to catch up. Thanks for your patience.

We’re investigating a problem, affecting many users, where messages sent via SMS to Twitter are not being received. 

There are also people who are not receiving updates from Twitter; we are working on this as well.

We’ve received notification from our SMS provider that a problem has been resolved that would have affected delivery for a large number of mobile users inside the US. A broken connection between our provider and the mobile operators caused a significant number of texts to go undelivered. This connection has since been fixed.

We are still investigating a number of lingering issues. We know that there are still cases where folks are not getting all the texts they should. So while we believe the situation should be improved, there is still more to do. 

Many Verizon users are still experiencing problems with SMS delivery. We are working closely with the providers we rely on to delivery these messages in order to get delivery working smoothly again.

We’re having SMS delivery problems to T-mobile and other operators. We’re tracking down these problems now and believe they primarily exist upstream from us.

This function will be back in all its glory tomorrow.

We’re working on problems affecting users sending and receiving text messages over the Canadian short code.