Ben McIlwain
dcfe939c38
Summarize all documentation updates across the repository to align with modern GKE, Cloud SQL Proxy v2, standard EPP fee v1.0, and Postgres database environments. Key Updates: - Prerequisites: Bump Java requirement to Java 25. - Architecture & Scaling: Document GKE workloads, Cloud Tasks queues, and scheduled tasks. Replace App Engine references with GKE deployment restart commands (kubectl rollout restart). - Configuration: Update Cloud SQL Proxy instructions to v2, fix keyring verification commands, and document IAP configuration. - Escrow (RDE/BRDA): Fix manual generation and download procedures to match the Dataflow job ID folder structure, and correct deposit encryption/verification command parameters. - Monitoring: Correct metric names and expand the documented metrics list with caching, locking, and reserved list metrics. - Fixes: Standardize lists formatting across markdown files, fix broken webdriver links, and resolve various typos. - Cleanup: Remove leftover cloud scheduler configurations for the deleted wipeOutContactHistoryPii task, and update ICANN reporting documentation to reflect open-sourced DNS query coordinator. TAG=agy CONV=88271e71-e272-40e0-85f8-a075a423b7c2
7.0 KiB
Operational procedures
This document covers procedures that are typically used when running a production registry system.
Stackdriver monitoring
Stackdriver Monitoring is used to instrument internal state within the Nomulus internal environment. This is broadly called white-box monitoring. EPP, DNS, and RDAP are instrumented. The metrics monitored are as follows:
/dns/publish_domain_requests-- A count of publish domain requests, described by the target TLD and the publish status./dns/publish_host_requests-- A count of publish host requests, described by the target TLD and the publish status./epp/requests-- A count of EPP requests, described by command name, client (registrar) id, and return status code./epp/request_time-- A Distribution representing the processing time for EPP requests, described by command name, traffic type, and return status code./rdap/requests-- A count of RDAP requests, described by endpoint type, deleted inclusion, registrar specification, authorization, and HTTP method./rdap/request_time-- A Distribution representing the processing time for RDAP requests, described by endpoint type, search type, wildcard type, HTTP status code, and incompleteness warning type./lock/acquire_lock_requests-- A count of lock acquisition attempts, described by TLD, resource name, and the existing lock state./lock/lock_duration-- A Distribution representing the lock lifetime in milliseconds, described by TLD and resource name./cache/lookups-- A count of cache lookups, described by cache name (e.g. domain, host) and the hit type (LOCAL, REMOTE, MISS, MISS_NONEXISTENT)./domain_label/reserved/checks-- A count of reserved list checks, described by TLD, number of matching lists, most severe list name, and most severe reservation type./domain_label/reserved/processing_time-- A Distribution representing the amount of time in milliseconds required to check a label against all reserved lists./domain_label/reserved/hits-- A count of reserved list hits, described by TLD, reserved list name, and the reservation type found.
Follow the guide to set up a Stackdriver account and associate it with the GCP project containing the Nomulus app. Once the two have been linked, monitoring will start automatically. For now, because the visualization of custom metrics in Stackdriver is embryronic, you can retrieve and visualize the collected metrics with a script, as described in the guide on Reading Time Series and the custom metric code sample.
In addition to the included white-box monitoring, black-box monitoring should be set up to exercise the functionality of the registry platform as a user would see it. This monitoring should, for example, create a new domain name every few minutes via EPP and then verify that the domain exists in DNS and RDAP. For now, no black-box monitoring implementation is provided with the Nomulus platform.
Updating cursors
In most cases, cursors will not advance if a task that utilizes a cursor fails (so that the task can be retried for that given timestamp). However, there are some cases where a cursor is updated at the end of a job that produces bad output (for example, RDE export), and in order to re-run a job, the cursor will need to be rolled back.
In rare cases it might be useful to roll a cursor forward if there is some bad data at a given time that prevents a task from completing successfully, and an acceptable solution is to simply skip the bad data.
Cursors can be updated as follows:
$ nomulus -e {ENVIRONMENT} update_cursors exampletld --type RDE_STAGING \
--timestamp 2016-09-01T00:00:00Z
Change cursorTime of RDE_STAGING for Scope:exampletld to 2016-09-01T00:00:00Z
Perform this command? (y/N): Y
Running ...
Updated 1 cursors.
gTLD reporting
gTLD registry operators are required by ICANN to provide various reports (ccTLDs are not generally subject to these requirements). The Nomulus system provides some of these reports, but others will need to be implemented using custom scripts.
Registry Data Escrow (RDE)
RDE is a daily deposit of the contents of the registry, sent to a third-party escrow provider. The details are contained in Specification 2 of the registry agreement.
Nomulus provides code to generate and send these deposits.
Monthly registry activity and transaction reporting
ICANN requires monthly activity and transaction reporting. The details are contained in Specification 3 of the registry agreement.
These reports are generated by querying BigQuery, using database snapshots
loaded into BigQuery. The default DnsCountQueryCoordinator implementation
(CloudDnsCountQueryCoordinator) relies on Google-internal DNS tables, so
external users will need to provide their own implementation to query their DNS
statistics.
Zone File Access (ZFA)
ICANN requires a mechanism for them to be able to retrieve DNS zone file information. The details are contained in part 2 of Specification 4 of the registry agreement.
This information will come from the DNS server, rather than Nomulus itself, so ZFA is not directly part of the Nomulus release.
Bulk Registration Data Access (BRDA)
BRDA is a weekly archive of the contents of the registry. The details are contained in part 3 of Specification 4 of the registry agreement.
ICANN uses sFTP to retrieve BRDA data from a server provided by the registry. Nomulus provides code to generate these deposits, but a separate sFTP server must be configured, and the deposits must be moved onto the server for access by ICANN.
Spec 11 reporting
Spec 11 reporting must be provided to ICANN as part of their anti-abuse efforts. This is covered in Specification 11 of the registry agreement, but the details are little spotty. Nomulus provides code to generate and send these reports, run on a schedule