Now Resolving: No tweets through twitter streaming api

davykestens · 2013-11-21T00:37:54.000Z

We’re seeing the same downtime here.

woriobot_a · 2013-11-21T00:38:32.000Z

We’re seeing the same thing. Frustrating that the API status page isn’t showing anything wrong.

LogicalArthur · 2013-11-21T00:43:05.000Z

Hi everyone–

Thank you for reporting this issue. We’re aware of it and are working to resolve it.

Some fixes may already have taken effect. Please feel free to share here if you’re able to connect reliably again, and please let folks know what hostname and endpoint you’re able to connect to.

talkingfrog1950 · 2013-11-21T00:44:38.000Z

We are down as well. For awhile we were logging stream closed as soon as we opened a user stream on restart. But now it is just sitting there. No 420 error warning of exceeding connections.

nyxtom · 2013-11-21T00:49:58.000Z

We are experiencing the same thing. Good to see that I’m not the only one.

nyxtom · 2013-11-21T00:52:09.000Z

We’re experiencing the same thing since about 4:45PM MDT.

talkingfrog1950 · 2013-11-21T01:03:47.000Z

UserStream service is still operating normally according to status page

https://dev.twitter.com/status

stu1011 · 2013-11-21T01:09:51.000Z

It’s back for me.

LogicalArthur · 2013-11-21T01:15:28.000Z

Hi again, everyone:

Endpoints on stream.twitter.com should be up and running (make sure to clear your DNS cache before trying to reconnect). However, success rates for endpoints on userstream.twitter.com and sitestream.twitter.com are still low.

Thanks for your patience.

woriobot_a · 2013-11-21T01:19:32.000Z

Just starting working for us again. Thanks!

meshfire · 2013-11-21T01:30:00.000Z

Thank you for the update, @LogicalArthur. We’ll keep an eye here for your further updates on userstream.twitter.com.

LogicalArthur · 2013-11-21T02:01:20.000Z

Update: userstream.twitter.com streaming endpoints should now be fully operational.

After flushing your DNS cache and attempting to reconnect, if you are still experiencing difficulties, please post here with the following:

Timestamp
Your IP address
The username under which your app is running
The IP address userstream.twitter.com resolves to on the machine you’re attempting connection from

bmaso · 2013-11-21T02:03:19.000Z

Just reporting I started receiving a trickle of tweets at about 2013-11-21 01:10 GMT.

(I was stuck underground in a subway so couldn’t report sooner. Leave it to LA to have a subway with no wifi and no cell reception possible. Nice to come out and see everything started working without my help!)

Brian Maso

LogicalArthur · 2013-11-21T02:10:23.000Z

Update: sitestreams.twitter.com streaming endpoints should now be fully operational as well.

That means that at this time, all streaming endpoints should be fully operational. If you continue to experience difficulty, please submit the information I posted above with respect to User Streams, but include the hostname and endpoint you’re trying to connect to as well.

We may be shifting traffic between our datacenters without notice, so please make sure that you are not caching DNS records for more than a few minutes at most, so as to allow us to more effectively loadbalance and make sure you’re able to connect.

Thanks for your patience, everyone, and sorry for the inconvenience.

meshfire · 2013-11-21T02:10:23.000Z

We appear to be running normally again. We’ll keep an eye on things and report if anything changes. Thanks again!

nyxtom · 2013-11-21T03:39:00.000Z

Thanks @LogicalArthur. I was curious why these uptime stats and updates to the stream weren’t reported to status.twitter.com or the API status. It’s often difficult to know right away without opening up a discussion forum. It also seems a bit misleading and hard to trust the https://dev.twitter.com/status when it shows apparently 100% uptime over the last 24 hours when in fact it was down for several hours for some of us today. Is there something that could help me understand why there is such a large discrepancy in the status being reported?

yegle · 2013-11-21T06:40:00.000Z

Hi @LogicalArthur

I’m still having trouble to access the streaming API.

I’m collecting data for academic research and I’m using https://stream.twitter.com/1.1/statuses/filter.json and https://stream.twitter.com/1.1/statuses/sample.json concurrently.

Now the strange thing is, when I just using one of the API to receive stream, there’s nothing wrong. But when I start the second script to receive stream, the first connection just dropped (TCP connection status change from ESTABLISHED to CLOSE_WAIT)

Timestamp: right now 1385015910
My IP address: 142.4.214.71
Username: yegle
IP adress stream.twitter.com resolve to: 199.59.148.229 and 199.59.148.138

LogicalArthur · 2013-11-21T08:16:00.000Z

@yegle:

Please note [node:6344, hash=“Connections”, title=“the documentation about simultaneous connections”]:

“Each account may create only one standing connection to the public endpoints, and connecting to a public stream more than once with the same account credentials will cause the oldest connection to be disconnected.”

yegle · 2013-11-21T08:47:54.000Z

Hi @LogicalArther

Thank you for the clarification. It’s strange since I’ve been using two public streaming API concurrently for 3 days. But anyway thank you for the help.

episod · 2013-11-21T15:56:04.000Z

The system that powers the status page is showing its age isn’t it? It’s a simple third party tool looking for simple HTTP response codes; in this case the conditions it was looking for were being met. Some day we hope to have a more tightly integrated automated status reporting system, but for today we have the current system which is better at reporting exceptional states (usually errors in the HTTP 50X series) for the REST API than the Streaming family of APIs.