Skip to content

Releases: tzConnectBerlin/que-pasa

v1.2.6: --main-schema feature fix

29 Aug 12:01
a8440bb
Compare
Choose a tag to compare

For more info see the related PR: #60

v1.2.5: point ithacanet->ghostnet, and add jakartanet support

30 Jun 14:50
5d4f423
Compare
Choose a tag to compare
Merge pull request #59 from tzConnectBerlin/ghostnet-and-jakartanet

point ithacanet->ghostnet, and add jakartanet support

v1.2.4

29 Jun 10:44
a56e737
Compare
Choose a tag to compare

Critical bugfix (see #58 for more info) - bug triggered by new Octez version that kicked in with the jakarta upgrade

v1.2.3

11 May 15:19
90fcdba
Compare
Choose a tag to compare

Critical enough bugfix for a new release. See details: #54.

v1.2.2: _live derived tables bugfix

09 May 17:07
Compare
Choose a tag to compare

When a multi-element datatype was erased (emptied), _live remained unupdated. This has been fixed now.

For more info, see the related PR: #53

V1.2.1: indexing of contract call parameters

04 May 10:34
Compare
Choose a tag to compare

Que Pasa now also indexes the parameters of contract calls, and writes them into new tables under entry.<entrypoint> (plus potential child tables, if the entrypoint has a datatype that folds into multiple tables).

For more details, see the related PR: #44

v1.1.4: add mode for only applying migrations

04 Apr 19:51
Compare
Choose a tag to compare
Merge branch 'main' of github.com:tzConnectBerlin/que-pasa

v1.1.3: allow retrying seemingly non-transient errors (like 404, 500, 400)

09 Mar 11:15
Compare
Choose a tag to compare

1). Added a CLI argument (--node-comm-retries) that allows to define number of retries on any kind of node communication error (that is not transient like a rate limit error, those are already retried by default).

This is useful for when a node url is used where multiple nodes are behind, each may be at different block heights which can result in a 404 from time to time.

2). Fixed a query that potentially was very slow in a specific corner case: when bootstrapping a contract with --bcd-enable, and this contract was created a long time ago, and this contract has had barely any activity. New query is practically instant, also in this corner case.

v1.1.2: 2 fixes related to bootstrap

03 Mar 15:17
7689ec1
Compare
Choose a tag to compare

For details see this PR: #46

v1.1.1: improved db and node communication issues recovery

22 Feb 12:25
Compare
Choose a tag to compare

A few new features for increased que pasa resilience to node and database communication issues:

  • Added application of a db pool library (r2d2), for managing retry on temporary connection issues logic.
  • Added possibility to specify multiple node urls, the first one being primary, the subsequent ones being fallback nodes in case of non-recoverable issues with the primary node.
  • Added curl code 56 to the list of transient node communication errors.
  • Don't retry empty body responses, because this does not necessarily mean it's a temporary issue with the node. It can also be due to an incorrect specified contract address, which will never succeed. Instead, replaced this with checking the http status code, and retrying when it is a rate limit issue (http code 429)

Unfortunately, adding r2d2 mend that we lost the capability of connecting to the database with --ssl natively. As a replacement we've added a little script in script/ that applies stunnel. If ssl is required to connect to the database, this script should be called now which will setup an ssl tunnel. Then que pasa can simply target the local end of this tunnel without ssl.