Traffic cannot flow error

Not applicable

I am consistently receiving the error below when deploying a proxy after rebuilding it several times.

The revision is deployed, but traffic cannot flow. Unexpected error occurred while processing the updates

What does this mean?bliaus-webservices-rev6-2015-04-21.zip

Solved Solved
1 8 3,242
1 ACCEPTED SOLUTION

You can also make a management call to check the deployments to find which router or message processor has an error for your deployment and the state of the deployment of the API proxy. For example:

Cloud -

$ curl -u example@apigee.com https://api.enterprise.apigee.com/v1/o/janicehunt/e/prod/deployments

{
{
  "aPIProxy" : [ {
  "name" : "weatherapi",
  "revision" : [ {
  "configuration" : {
  "basePath" : "/",
  "steps" : [ ]
  },
  "name" : "2",
  "server" : [ {
  "status" : "deployed",
  "type" : [ "message-processor" ],
  "uUID" : "e6f83b53-8ab1-4325-bb77-df1bc4df8d59"
  }, {
  "status" : "deployed",
  "type" : [ "message-processor" ],
  "uUID" : "c30cd253-63ec-4c35-905e-5c78c51bb0e4"
  }, {
  "status" : "deployed",
  "type" : [ "router" ],
  "uUID" : "3341eda0-fa82-4dcb-bbf1-5cf77dc03efc"
  }, {
  "status" : "deployed",
  "type" : [ "router" ],
  "uUID" : "c0b782f2-ccb6-4984-a0c0-0ac1db60a039"
  }, {
  "status" : "deployed",
  "type" : [ "router" ],
  "uUID" : "38f80c53-a22b-4427-ab95-a178482daa6b"
  }, {
  "status" : "deployed",
  "type" : [ "router" ],
  "uUID" : "28d847a2-5bfd-404d-985d-33d079d691d9"
  }, {
  "status" : "deployed",
  "type" : [ "router" ],
  "uUID" : "26467f86-e529-4135-b012-120e274755d3"
  }, {
  "status" : "deployed",
  "type" : [ "router" ],
  "uUID" : "807be0c7-dcc7-4be4-9d1e-30d26fa5519c"
  } ],
  "state" : "deployed"
  } ]
  },

From the deployment output, check for the "status" : "error" on any router or message processor. If you are using the Apigee Private Cloud (OPDK), go to the router or message processor node identified by the uuid and check the /<install-root>/apigee4/var/log/apigee/{router|message-processor}/log/system.log file and see if you can find any exceptions related to why the deployment failed. Most of the time it might be a connection level problem to Cassandra datastore which may have contributed to the problem. Typically a restart of that message processor or router node should help in properly deploying the bundle.

View solution in original post

8 REPLIES 8

Not applicable

@Terry David

Where are you seeing this error ? This happens when your management servers are unable to reach router and message-processors and time out finally .

If this OPDK probably you can check that in logs . this is not related to your bundle for sure .

You can also make a management call to check the deployments to find which router or message processor has an error for your deployment and the state of the deployment of the API proxy. For example:

Cloud -

$ curl -u example@apigee.com https://api.enterprise.apigee.com/v1/o/janicehunt/e/prod/deployments

{
{
  "aPIProxy" : [ {
  "name" : "weatherapi",
  "revision" : [ {
  "configuration" : {
  "basePath" : "/",
  "steps" : [ ]
  },
  "name" : "2",
  "server" : [ {
  "status" : "deployed",
  "type" : [ "message-processor" ],
  "uUID" : "e6f83b53-8ab1-4325-bb77-df1bc4df8d59"
  }, {
  "status" : "deployed",
  "type" : [ "message-processor" ],
  "uUID" : "c30cd253-63ec-4c35-905e-5c78c51bb0e4"
  }, {
  "status" : "deployed",
  "type" : [ "router" ],
  "uUID" : "3341eda0-fa82-4dcb-bbf1-5cf77dc03efc"
  }, {
  "status" : "deployed",
  "type" : [ "router" ],
  "uUID" : "c0b782f2-ccb6-4984-a0c0-0ac1db60a039"
  }, {
  "status" : "deployed",
  "type" : [ "router" ],
  "uUID" : "38f80c53-a22b-4427-ab95-a178482daa6b"
  }, {
  "status" : "deployed",
  "type" : [ "router" ],
  "uUID" : "28d847a2-5bfd-404d-985d-33d079d691d9"
  }, {
  "status" : "deployed",
  "type" : [ "router" ],
  "uUID" : "26467f86-e529-4135-b012-120e274755d3"
  }, {
  "status" : "deployed",
  "type" : [ "router" ],
  "uUID" : "807be0c7-dcc7-4be4-9d1e-30d26fa5519c"
  } ],
  "state" : "deployed"
  } ]
  },

From the deployment output, check for the "status" : "error" on any router or message processor. If you are using the Apigee Private Cloud (OPDK), go to the router or message processor node identified by the uuid and check the /<install-root>/apigee4/var/log/apigee/{router|message-processor}/log/system.log file and see if you can find any exceptions related to why the deployment failed. Most of the time it might be a connection level problem to Cassandra datastore which may have contributed to the problem. Typically a restart of that message processor or router node should help in properly deploying the bundle.

adas
Participant V

@Terry David

Like Janice pointed out, you might want to make the management call to get the deployment status of your api for a given revision. It would list down the status of the deployment for each of your routers and MPs. Typically this happens if one of your MP or routers hasn't come up properly or some service didn't get started properly due to a startup error. This can also happen incase the RPC port between your router/mp and management server is not open or the connection is too slow such that the call times out. You would find the corresponding error in the system.logs of your management server or router/mp.

Not applicable

@Arghya das: when working in the Apigee cloud, is there anyway to resolve that error?

Wondering the same thing as @Sven Huberti. I have a eval cloud account and am jammed up. Any way to get unstuck? Would rather not have to recreate my policy 😞

Hi @Travis Spencer,

We've recently been working on a set of troubleshooting topics. Please take a look at this one. If you look in the TOC on the left side of the doc page, you'll see a few other similar errors listed -- find the one that matches your situation. The advice given is similar to the advice given here, but with some more context.

As mentioned above, it's possible you will not be able to recover from this error without help from Apigee after trying the troubleshooting steps listed. If you can't get unstuck, please post the information listed here to the community.

Will

Thanks for the link to the troubleshooting guide, @wwitman. I went through those, but couldn't unjam it myself 😞 So, I created a new question on the forum.

adas
Participant V

@Sven Huberti Unfortunately no.

Depending on what the root cause is you may or may not be able to recover from this error on your own in the cloud. For example if the deployment failed to one of the MPs or routers due to rpc timeout, you might want to do a force undeploy (with query param force=true) and redeploy which would fix the problem. If it was a startup issue with the MP and doesn't get resolved even after redeploy, you would have to contact support to have these instances restarted.