Increase the XRAY cache size

[Jose-Matsuda]

Possibly related to #1129

When the XRAY data folder gets too full the indexer "stops" working in that it will not scan any new packages

Example under pypi-remote-cache/35/ad/616c27cade647c2a1513343c72c095146cf3e7a72ace6582574a334fb525/pandas-1.4.2-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl

Compared to a one that was scanned

The cache itself is cleaned out on a regular basis, but if there are a lot of packages being scanned then its possible for the cache to become too full.

Need to find out where and how to increase the cache size for XRAY, we may want to really increase the size here especially if AVD's end up using it.

[Jose-Matsuda]

Recall that we install the jfrog platform using this chart

Looking into our prod jfrog.tf I find the following

xray:
  enabled: true
  database:
    user: xray@...
    actualUsername: xray
    password: '...'

And then looking at the documentation here I see the following under database

## Database Configuration
  ## PostgreSQL database used by all xray services
  database:
    #type: postgresql
    #driver: org.postgresql.Driver
    #url: "postgres://<your db url, for example: localhost:5432>/xraydb?sslmode=disable"
    #username: xray
    #password: xray
    ## Some managed databases like Azure have a different username for accessing the database than the actual one inside the database
    ## for Azure managed postgres for example username will be xray@mycompany and the actualUsername will be xray
    #actualUsername: ""

so it does seem like the database portion of this is seemingly out of our hands, or at least I don't know how to increase the size there.

[Jose-Matsuda]

CLOUD ticket 14321

[Jose-Matsuda]

It’s not that it isn’t managed there (in reference to looking at the jfrog tf file), it’s just the PVC already exists

1. Scale down xray statefulset
2. Change PVC size with kubectl edit (edited)
3. Scale back up

Online resizing is not available today. I believe there is a preview feature but we don’t have it

Make sure you wait until the pvc says waiting for pod restart before scaling back up

https://carlos.mendible.com/2021/11/15/aks-resize-private-volume-claim-to-expand-a-managed-premium-disk/

[Jose-Matsuda]

Finished courtesy of Chuck