Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Mainnet #2

Open
wants to merge 19 commits into
base: master
Choose a base branch
from
Open

Mainnet #2

wants to merge 19 commits into from

Conversation

gisele-igore
Copy link

No description provided.

harryttd and others added 19 commits January 12, 2023 16:42
* Exit config-init script on error

* Update chart tests
A bug was uncovered during mumbai activation, where nodes were unable to
activate and were crashing.

The [fix](https://gitlab.com/tezos/tezos/-/merge_requests/7462/diffs)
broke dailynet because it is no longer possible to activate with a
fitness of -1 (which is how we have been activating chains since the
beginning of tezos-k8s).

Replacing this value to 1 fixes the issue for dailynet and also does not
appear to be causing any issue with v15.1.
* tezos-proto-cruncher chart

This chart deploys a daemonset in your cluster. the daemonset runs a
parallel search for a vanity protocol hash.

It takes s3 bucket credentials. Protocol is downloaded from bucket then
results are written to the bucket.

It does not terminate: it keeps running until the cluster is torn down.

* better variable names

* Update charts/tezos-proto-cruncher/scripts/proto-downloader.py

Co-authored-by: Aryeh Harris <[email protected]>

* address review comments

* lint python scripts with black

* catch exception when s3 upload fails

* fix manual method

* fix manual method, better way per Aryeh

* fail fast when env vars not set in proto downloader

Co-authored-by: Aryeh Harris <[email protected]>
* Update bcd indexer

* Add storageClassName field

* Make bcd images configurable

* black .

* Update BCD indexer to 4.4.0

---------

Co-authored-by: Nicolas Ochem <[email protected]>
…a#530)

* change tezos_version information in metadata

* new metadata format version compare

* schema validation

* add pip package and better log for json schema validate

* black formatting

* typos and better function error handling

* check-jsonschema needs python headers for ruamel

* fixes:
--network should always get chain_name
typo with rpc info object

* rm --in memory it was retired

* rm network version

* strings to ints

* skip old metadata style

* typos

* clean up metadata

* typo

* move snapshot header cat

* conditional --json for v16 and v15

* nicolas version correction

* fix sh var quoting

* Get TEZOS_VERSION before you check it

* Fix version assignment

* fix assignments

* new version string

* more graceful handling of pre-schema artifacts

The exception handling of pre-schema metadata was too broad, and would
prevent the webpage from being generated at all unless all three types
are present - including archive.

* fix UI bug - display more recent artifact instead of oldest

* Rm erroneous context_elements using commit_hash

* Move rpc info out of each type since its universal

* Overwrite release object in metadata

* Don't upload empty metadata files

* Rm context elements and add --json to snapshot info

* Validate tezos-snapshots.json isntead of individual files

* add default schema url

* validation deps and failure if no validation

* replace version with $schema url

* cleanup

* Fix imports

* Need another level for new metadata format

* update default value per nicolas request

* config generator needs to know where snapshots are

* https default snapshot engine schemaurl

---------

Co-authored-by: Nicolas Ochem <[email protected]>
* snapshot metadata resiliency + new format

* support new metadata format where the top-level object is a dict
* support new version format matching
* do not even attempt to download or parse metadata if the context
  folder is already here (this prevents impact of xtz-shots downtime on
  preexisting nodes)
* when we actually attempt to download a snapshot, do not fail
  silently, always raise. (this will cause the pod to restart until
  xtz-shots is fixed, then will download normally)

* fix octez version match
Do not fetch snapshot metadata when data is already present - fix typo
in previous merge.
* Network argument is broken so we're removing it

`--network` has been retired as of v16 and it is redundant to `--data-dir`. We've tested it on both v15 and v16 and the omission of this does not break backwards compatability.

* regen tests
* Working BCD frontend

* Expose port 8080 on bcd service for gui

* update sandbox config and mount to indexer container

* Make pg volume storage configurable

* update sandbox config

* require rpcUrl and apiUrl to be set for gui

* Clean up prop access

* Optionally create a postgres db container + make DB_HOSTNAME a secret

* Update error msg

* Make postgres image configurable

* b64enc hostname

* Only set PGDATA if using internal postgres db

* bug fix
…a#554)

With v15 to v16 upgrades, we started seeing failures updating storage of
non-mainnet networks.

We reported upstream:

https://gitlab.com/tezos/tezos/-/issues/4892#note_1287844009

However it turns out that the data dir we were passing did not have a
config file, so it was unable to figure out which network the data dir
was for.

This is fixed by passing config file instead of data dir.

I tried by launching v15 ghostnet from a snapshot, then upgrading to
v16. I also tried launching a v16 tarball directly from scratch. Both
worked.

test
* rename tezos-baker to octez-baker

now that v14 is no longer relevant we can finish renaming. The current
terminology will probably be deprecated in v17.

* fix octez-signer as well per review
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

5 participants