ec2nommered stopped working

[Arametheus]

Hello, I emailed you also but figure post here just incase.

we had everything set-up and sending testing data using test buckets we have. I needed to make a config change for prod and relaunched feederd and everything seemed fine.

I send my job request to feederd Instance and it excepts it like normal. but here is were the issue is coming in.

the jobs in simple DB media_nommer stay in pending state and never get processed. I checked the SQS messages and looks like feedered is sending messages.. but doesn't look like the ec2nommer instance is picking anything up and processing them.

I tried switching back to our test buckets since that was working just incase of some kind of security but that no longer works so I'm really confused.

any help would be great.

[Arametheus]

when I logged into the nommer just to see this is in the log
Unhandled error in Deferred:
Unhandled Error
Traceback (most recent call last):
File "/home/nom/.virtualenvs/media_nommer/local/lib/python2.7/site-packages/twisted/plugins/ec2nommerd.py", line 108, in start_tasks
interval_tasks.register_tasks()
File "/home/nom/.virtualenvs/media_nommer/local/lib/python2.7/site-packages/media_nommer/ec2nommerd/interval_tasks.py", line 92, in register_tasks
now=True)
File "/home/nom/.virtualenvs/media_nommer/local/lib/python2.7/site-packages/twisted/internet/task.py", line 163, in start
self()
File "/home/nom/.virtualenvs/media_nommer/local/lib/python2.7/site-packages/twisted/internet/task.py", line 194, in call
d = defer.maybeDeferred(self.f, _self.a, *_self.kw)
--- ---
File "/home/nom/.virtualenvs/media_nommer/local/lib/python2.7/site-packages/twisted/internet/defer.py", line 133, in maybeDeferred
result = f(_args, *_kw)
File "/home/nom/.virtualenvs/media_nommer/local/lib/python2.7/site-packages/media_nommer/ec2nommerd/interval_tasks.py", line 49, in task_check_for_new_jobs
jobs = JobStateBackend.pop_new_jobs_from_queue(num_jobs_to_pop)
File "/home/nom/.virtualenvs/media_nommer/local/lib/python2.7/site-packages/media_nommer/core/job_state_backend.py", line 477, in pop_new_jobs_from_queue
visibility_timeout=3600)
File "/home/nom/.virtualenvs/media_nommer/local/lib/python2.7/site-packages/media_nommer/core/job_state_backend.py", line 433, in _pop_jobs_from_queue
raise Exception(msg)
exceptions.Exception: SQS only allows up to 10 messages to be popped at a time.
nommerd nomconf S3 URI present, using that to download nomconf.py
2014-08-02 20:40:38+0000 [-] Log opened.
2014-08-02 20:40:38+0000 [-] twistd 11.0.0 (/home/nom/.virtualenvs/media_nommer/bin/python 2.7.2) starting up.
2014-08-02 20:40:38+0000 [-] reactor class: twisted.internet.selectreactor.SelectReactor.
2014-08-02 20:40:38+0000 [-] twisted.web.server.Site starting on 8002
2014-08-02 20:40:38+0000 [-] Starting factory <twisted.web.server.Site instance at 0x248e050>
~
guessing something is wrong with the nommer ami.. but not sure.

[gtaylor]

Hello,

I wish I could be of more help, but I don't see anything obviously wrong. The AMI hasn't changed in close to two years now, so I'm not too sure what could be causing your issues.

I need to update the documentation to reflect this, but we have more or less stopped using media-nommer after AWS released ElasticTranscoder. They have a dedicated team whose job is to work on encoding. We were working on media-nommer on the side, it wasn't even a core part of our business. ElasticTranscoder is going to be more reliable, faster, will support more formats, and is even cheaper than media-nommer at certain scales. There aren't too many things that media-nommer does better anymore.

I realize that this may not be the answer you are looking for, but you are probably better off with ET unless you are doing some deep customizations to support special encoding workflows.

[Arametheus]

Greg. Np yea it was weird since it was working fine. But I was working on switching to ET anyways cause of the same reasons you mentioned. It Just sped up my timeframe to do it.

Sent from my iPhone

On Aug 3, 2014, at 2:19 PM, Greg Taylor [email protected] wrote:

Hello,

I wish I could be of more help, but I don't see anything obviously wrong. The AMI hasn't changed in close to two years now, so I'm not too sure what could be causing your issues.

I need to update the documentation to reflect this, but we have more or less stopped using media-nommer after AWS released ElasticTranscoder. They have a dedicated team whose job is to work on encoding. We were working on media-nommer on the side, it wasn't even a core part of our business. ElasticTranscoder is going to be more reliable, faster, will support more formats, and is even cheaper than media-nommer at certain scales. There aren't too many things that media-nommer does better anymore.

I realize that this may not be the answer you are looking for, but you are probably better off with ET unless you are doing some deep customizations to support special encoding workflows.

—
Reply to this email directly or view it on GitHub.