Skip to main content

​
Migrate to the Cloud portal

​
For users on EVM and Substrate networks

Portal is currently in beta. Please report any bugs or suggestions to the SQD Portal Beta chat or to Squid Devs.
This guide walks you through replacing a gateway of the open private version of SQD Network with a portal of the permissionless SQD Network as the primary source of data for your Cloud squids. Benefits include:
  • Reduced reliance on centralized services: The permissionless SQD Network consists of over 2500 nodes ran by independent operators
  • Improved stability: With a total capacity of roughly 2Pb, the permissionless SQD Network allows for a great deal of redundancy.
  • Improved speed: The permissionless version of SQD Network has a lot more bandwidth than the open private network; moreover, portals use the available bandwidth more effectively than gateways. Data fetching was 5-10 times faster in some of our tests. For squids that are not bottlenecked by write operations this will translate into better sync performance.
  • Being future-proof: All future development will be focused on portals and the permissionless SQD Network. —> SQD team operates two independent portals that serve the needs of Cloud users:
  • The dedicated Cloud Portal is only visible from within the Cloud, ensuring stable performance for squids deployed there.
  • The Public Portal can be accessed from anywhere for easy experimentation and local development.
We’re currently experimenting with tightening the data request complexity limits. If you see an HTTP 400 error with a message like this: Couldn't parse query: query contains X item requests, but only 50 is allowed where X is some number above 50, or any other HTTP 400 response, please let us know.
Here are the steps to migrate:

​
Step 1: update to @latest packages

A. Enter your squid’s folder. B. Remove both your lock file and the node_modules folder: 
<Tabs queryString="package-manager">
<Tab value="npm" label="NPM">
```
```bash
rm -r node_modules package-lock.json

</Tab>
<Tab value="yarn" label="Yarn">
```
```bash
rm -r node_modules yarn.lock

</Tab>
<Tab value="pnpm" label="PNPM">
```
```bash
rm -r node_modules pnpm-lock.yaml

</Tab>
</Tabs>
```

**C.** Upgrade all SQD packages to the `@latest` major version:
```bash
npx --yes npm-check-updates --filter "@subsquid/*" --upgrade
```

**D.** Install the dependencies:

```mdx-code-block
<Tabs queryString="package-manager">
<Tab value="npm" label="NPM">
```
```bash
npm install

</Tab>
<Tab value="yarn" label="Yarn">
```
```bash
yarn install

</Tab>
<Tab value="pnpm" label="PNPM">
```
```bash
pnpm install

</Tab>
</Tabs>

​
Step 2: tell your code to use the Portal

Follow the network-specific instructions from the Cloud: Once you’re done, your squid will use the Cloud Portal when deployed and the public portal for local runs.

​
Step 3 (recommended): testing

If you just want to update your squid code to use the Portal-enabled SDK version you can simply redeploy your squid here and be done. No need to re-sync. However, for the duration of this beta we strongly recommend that you re-sync your squid. This will allow you to make sure everything is working as it should and evaluate the improved data source performance. Follow the zero-downtime update procedure:
  1. Deploy your squid into a new slot.
  2. Wait for it to sync, observing the improved data fetching.
  3. Assign your production tag to the new deployment to redirect the GraphQL requests there.
See this section for details.

​
What’s next?

Stay hyped for news on the introduction of real time data to EVM and Substate - it is coming soon.