8

I recently upgraded my azure storage libraries from 1.7 to 2.1 and since then I have not been able to deploy to my cloud service. I keep getting this deployment error.

The size of local resources cannot be reduced. Affected local resource is DiagnosticStore in WorkerRole. Current size is 20000. Requested size is 4096.

I haven't enabled caching and I tried using Small and even Medium VM's but still the issue persists. Any idea where the issue might be?

Bitsian
  • 2,238
  • 5
  • 37
  • 72

1 Answers1

11

Don't deploy directly to the "Production" slot.

Deploy to the "Staging" slot (which will requisition new resoruces/sizes) and then perform a VIP Swap.

Eoin Campbell
  • 43,500
  • 17
  • 101
  • 157
  • 1
    @Eoin ... I'm curious about why it would work considering staging and production environment are supposed to be identical. Can you please explain? Thanks. – Gaurav Mantri Sep 10 '13 at 15:48
  • staging and production environments are really just a logical construct at the RDFE (ie. portal/service management API) level. And swapping between the slots is only a config change at the load balancer. The two deployments themselves have nothing to do with each other. – kwill Sep 10 '13 at 20:31
  • 2
    @Guarav - Think about it like this. If you deploy once to "Production", it will put aside specific resources for that specific deployment. If you do an in-place redeploy, then those resources are already allocated, and all your doing is replacing the binaries/configs, hence you end up with a mis-match. When you deploy to staging, you are asking Azure to "Create me a completely new deployment, with new allocated resources, based on the config/binaries I'm about to upload". VIP Swapping is then just the Virtual IPs being repointed from one deployment to the other at the load balancer. – Eoin Campbell Sep 11 '13 at 08:13