Ask Your Question

Swift Proxy Account GET with 503 error

asked 2013-05-23 04:59:21 -0500

anonymous user


updated 2013-06-08 22:27:57 -0500

fifieldt gravatar image

Swift Proxy server throws Account GET failed with 503 internal server error. I am following the hastexo web site and have created storage nodes.The proxy-server.conf is as provided in site Please find the details below:

The command I have run is swift -v -V 2.0 -A -U admin:admin -K hastexo list The error I get is listed below. Account GET failed: 503 Internal Server Error

edit retag flag offensive close merge delete

3 answers

Sort by ยป oldest newest most voted

answered 2013-05-27 08:55:04 -0500

annegentle gravatar image

Did your swift services start correctly? Check for errors on startup and troubleshoot there, and also double-triple-check your /etc/swift/account-server.conf.

Since it looks like you are following the Hastexo guide for Essex, it's possible that some thing is different between your system and their recommendations because 12.04 also allows you to use the Ubuntu Cloud Archive for packages, which would give you Grizzly, and I'm not sure the instructions would exactly match. Can you indicate what OS you are installing on and whether you used the Cloud Archive?

edit flag offensive delete link more


I have Ubuntu 12.04 Precise Pangolin installed and I have a separate Openstack installed, then Swift Proxy Server and 3 storage nodes. However when I try to test. I am getting teh Account Get failed 503 internal server error. I have checked the /etc/swift/account-server.conf thats fine. Howevere when I try to start the nodes the /etc/swift/proxy-server does not start except that other services start when I try the command swift-init start all your help is appreciated

raghavendra gravatar imageraghavendra ( 2013-05-28 09:51:38 -0500 )edit

Does your Identity (keystone) installation work on the separate OpenStack? Or are you using a different middleware such as swauth? Also are you using the Cloud Archive or installing (Folsom) using 12.04? Swift logs all events in /var/log/syslog, so that is a good first place to look for errors.

annegentle gravatar imageannegentle ( 2013-05-28 13:03:06 -0500 )edit

Yes. I have a separate (Keystone) installation on the separate Openstack. The swift authentication in the Proxy server steps is provided below. I am installing Essex version of 12.04 Precise Pangolin and I have checked error log in /var/log/syslog. [DEFAULT] bindip = bindport = 8080 backlog = 4096 swiftdir = /etc/swift workers = 8 user = swift expiringobjectscontainerdivisor = 86400

You can specify default log routing here if you want:

logname = swift logfacility = LOGLOCAL0 loglevel = INFO

[pipeline:main] pipeline = catch_errors healthcheck cache authtoken swiftauth proxy-server

[app:proxy-server] use = egg:swift#proxy allowaccountmanagement = true accountautocreate = true set logname = swift-proxy-server set logfacility = LOGLOCAL0 set loglevel = DEBUG set accesslogname = swift-proxy-server set accesslogfacility = LOGLOCAL0 set accessloglevel = DEBUG set log_headers = True

[filter:healthcheck] use = egg:swift#healthcheck

[filter:catcherrors ...(more)

raghavendra gravatar imageraghavendra ( 2013-05-29 00:45:47 -0500 )edit

I also see that Since Ubuntu 12.04 LTS has OpenStack Essex by default, we can use Cloud Archives for Folsom /Grizzly. However my setup is with Essex Ubuntu 12.04. I do not get any errors when I run the commands for storage nodes, thats working fine.

raghavendra gravatar imageraghavendra ( 2013-05-29 06:17:11 -0500 )edit

I have the output of the storage node root@Swiftnode1:/# swift-init start all Starting container-updater...(/etc/swift/container-server.conf) account-auditor running (3254 - /etc/swift/account-server.conf) account-auditor already started... object-replicator running (3255 - /etc/swift/object-server.conf) object-replicator already started... Starting proxy-server...(/etc/swift/proxy-server.conf) swift-proxy-server does not exist container-replicator running (3257 - /etc/swift/container-server.conf) container-replicator already started... object-auditor running (3258 - /etc/swift/object-server.conf) object-auditor already started... object-expirer running (3347 - /etc/swift/object-expirer.conf) object-expirer already started... container-auditor running (3259 - /etc/swift/container-server.conf) container-auditor already started... container-server running (3261 - /etc/swift/container-server.conf) container-server already started... account-server running (3263 - /etc/swift/account-server.conf) account-server already started... Starting account-reaper...(/etc/swift/account-server.conf) container-sync running (3334 - /etc/swift/container-server.conf) container-sync already started... account-replicator running (3271 - /etc/swift/account-server.conf) account-replicator already started... Starting object-updater...(/etc/swift/object-server.conf) object-server running (3274 - /etc/swift/object-server.conf) object-server already ...(more)

raghavendra gravatar imageraghavendra ( 2013-05-29 06:41:15 -0500 )edit

answered 2015-12-12 13:54:50 -0500 (


Modify files above on storage nodes to have

pipeline = recon account-server

and append

use = egg:swift#recon
edit flag offensive delete link more

answered 2013-05-28 14:24:21 -0500

creiht gravatar image

I'm not familiar with the Hastexo instructions, but the logs should reveal the issue. I would start taking a look at the proxy server logs to see why a 503 was returned (and possibly which account servers are having issues), and then look at the account server logs to figure out what specific error is causing the issue. Each request returns a transaction id, which is logged with each request to make it easier to search for log lines that deal with that specific request.

edit flag offensive delete link more

Get to know Ask OpenStack

Resources for moderators

Question Tools


Asked: 2013-05-23 04:59:21 -0500

Seen: 6,677 times

Last updated: May 28 '13