Revisions of google-cloud-sap-agent
Robert Schweikert (rjschwei)
accepted
request 1208880
from
John Paul Adrian Glaubitz (glaubitz)
(revision 15)
- Update to version 3.6 (bsc#1231426, bsc#1231758) * Delete stale instant snapshot groups * Add collection for WLM Pacemaker metric `gcpstonith`. * Increase max retries for instant snapshot group to 8 from 3. * Fix fatih/color github package * PrintServiceStatus added to statushelper * Update generated proto files using protoc-gen-go v1.35.1. * adding set defaults method in instancemetadata * Fix ASCS discovery to handle commented lines and non-hostname values * Improve command used for HANA Version discovery * Add mutex to uap.go to prevent race condition. * Status OTE proto * Fix metrics sending for Backint * Update version OTE to print the version in the same format used by RPM * Add debug logs for HANA version discovery * Add a flag to indicate if an instance is part of a DR site. * Minor deleteISG changes * Add max retries to ISG service for exponential backoffs * update disk restore command param names and usage string * Change send-status-to-monitoring to send-metrics-to-monitoring to match flag name in usage string * Add additional error string to help debugging when metadata server call fails * Changing default naming of group snapshot to match mentioned format in flags. * Add usage examples for hanadiskbackup and hanadiskrestore * Add a log message to indicate that the HANA snapshot has been marked as SUCCESSFUL. * Implement module skeleton file layout for status helper for status check OTE * Add token scope for WIF authentication for ISG APIs * Lower the log level from INFO to DEBUG for messages generated by cloudmonitoring.sendBatch() to reduce log spam and prevent /var/log from filling up. * Add pagination support for instant snapshots * adding snapshot state metric in HANA Monitoring
Robert Schweikert (rjschwei)
accepted
request 1196063
from
John Paul Adrian Glaubitz (glaubitz)
(revision 14)
- Update to version 3.5 (bsc#1229668, bsc#1229669) * Add instance properties to all components based on instance role. * Refactor OTEs hanadiskbackup and supportbundle to enable different execution params for internal and user invocations - 2/n. * Backport the configureinstance increased timeout to 3.4 and 3.3 * Perform general X4 configurations first, then OS specific * Log the copy error only to log file and not to console * Add unit test to cover `gce.New` in systemdiscovery. * Adding IFNULL to log backup queries * Add HANA Compute Resources Report to performancediagnostics bundle. * Add mutex for backint event annotation logging * Only wait for shutdown in the initial start routine. Continue on during restarts. * Update backint log message for event annotations * Create HANA compute metrics report to add it to performancediagnostics. * Correct SAP System proto merging. * Change SAP_BACKINT_FILE_TRANSFER invocation * Adding proto for instance metadata OTE for wlm service * Report host maintenance events to Cloud Monitoring as part of ProcessMetrics data collection. Collection is done by querying the local metadata server. https://cloud.google.com/compute/docs/instances/host-maintenance-overview * Remove flaky unit test for canceled context in performancediagnostics. * Update startdaemon so LogRunningDaily and waitForShutdown behave as expected during configuration restarts. * Improve detection of Netweaver role to include SID check for multi-instance nodes. * App servers are no longer responsible for reporting DB scale type * Refactor computeresources to collect time-series disk IOPS usage metrics for performancediagnostics. * Fix typo in service_linux.go * Add SAP Workload Actions for start, stop, snooze. * Refactor OTEs hanadiskbackup and supportbundle to enable different
John Paul Adrian Glaubitz (glaubitz)
accepted
request 1183832
from
John Paul Adrian Glaubitz (glaubitz)
(revision 13)
- Update to version 3.4 (bsc#1227134, bsc#1227135) * Adding project to exclusion list * Add machine type to configure instance proto for WLM metric collection. * Add test channel for Guest Actions. Make default channel the registered channel. * Set backup object's customTime field as part of backint backups * Add workload discovery to configure command * Add multiple workers support in parallelreader for parallel downloading during restore. * `configureinstance` with `overrideVersion` set should log a warning and continue. * Minor log change in balanceirq * Add common function to parse parameters for guest action handlers * BalanceIRQ OTE added to Agent for SAP * Remove output from stdout for DIAGNOSE * Small hyperThreading change for configureinstance * Add initial steps to initialize the SystemDiscovery OTE in IIOTE and command mode. * Adding single worker support in parallelreader for download. * Read encryption key from file if specified in parameters file * Run configureinstance OTE only on supported instances during WLM metric collection. * Add instance ID to user agent string for SAP Agent. * Return `UsageError` as exit status instead of `Failure` in case of invalid parameters * Bumping up the agent version * Use json marshalling instead of manually parsing from map in configure handler * Move metric override modules to metricoverrides.go for general use * Updating the gcbdr proto * Updating param names to make it more clear in performance diagnostics * Add DiskSizeGb to Disk for disk creation. * Add Demo Metrics for Process Metrics * Add warning message for configureinstance overrideVersion * Add 3.3 to configureinstance versioning * Fix log message in configureinstance * Rename scope and param file to type and backint-param-file to avoid confusion
Robert Schweikert (rjschwei)
accepted
request 1177860
from
John Paul Adrian Glaubitz (glaubitz)
(revision 12)
- Update to version 3.3 (bsc#1225166, bsc#1225558) * Build updated to use -mod=vendor during build * Updated go.mod and go.sum with dependencies for safetext, using go mod vendor for github action * Add actual values and comments to usagemetrics.go to ensure that error and action codes are only appended to the end of the list. * Remove usage metrics from configureinstance.go * Add a hard Disable for reliability metrics collection until the namespace is created and tested. * Adding metrics for time taken by each query * Add SHA224 of labels as a new label. * Remove collect_reliability_metrics from configuration.json * Small tweaks to backint log and inquire path generation * Fix for unmarshalling backint configuration. * Implementation of instant snapshot group backup workflow * Backint changes around shorten_folder_path * Rename max_diagnose_size_gb to diagnose_file_max_size_gb * Adding start and finish logs in performance diagnostics * Validate that all disks mapped to /hana/data belong to the same consistency group. * Rename backint monitoring metrics parameter * Trim folder prefix for Backint INQUIRE output. * Add the ability to test the database connection * Reduce log level of some storage messages to debug. * Finalize guest action request and response format. * Backint dashboard fix logs * Add scorecards to backint dashboard * Making proto changes for HANA Monitoring support for multiple tennats and ha setup * Add total upload/download time to log. * Add HANA indexserver.ini metrics to WLM metric collection.
Robert Schweikert (rjschwei)
accepted
request 1164346
from
John Paul Adrian Glaubitz (glaubitz)
(revision 11)
- Update to version 3.2 (bsc#1222215, bsc#1222216) * Remove internal gensupport package. * Restore additional error handling and response checking to internal data warehouse client. * Updating the aggregate function in HANA insight rules * Remove a leftover debug log * Allow multipart uploads for PIPE file types. * Update go-hdb version to v1.8.0 * Perform log restores in serial rather than parallel. * Add sample usage examples to commandlineexecutor * Small update to configureinstance OTE * Add nil check in backup and restore flows to protect against panics. * Close http response body in WriteInsight() and soap.go * Record topology type. * Initialize usagemetrics for OTEs * Add Instance Number to SAP System instance properties * Set `min_version` for WLM `os_settings` system metric. * Increase timeout for saptune re-apply commands. * Adding handling for encrypted snapshots in backup and restore * Change the version check comparisons to account for versions older than those listed in SAP Note. * Skip the Netweaver metrics that need dpmon on NW kernels affected by SAP Note: 3366597 * Fix imports * No public description * Use internal data warehouse client. * Fix disp+work command invocation for Netweaver Kernel version discovery. * Add note about default parameter values to installbackint. * Add mutex in multipart writer for potential data races. * Update go.mod and go.sum * Skip XFS freeze by default unless user passes a parameter to do it explicitly
Robert Schweikert (rjschwei)
accepted
request 1148691
from
John Paul Adrian Glaubitz (glaubitz)
(revision 10)
- Update to version 3.1 (bsc#1220010, bsc#1220111) * Fixing system replication status code being returned * Reduce disk snapshot wait durations * Fix test flakes in workloadcollector test. * adding metrics for db freeze time and total workflow time * Fix for SAP System discovery adding the current host to all components. * Restore default WLM metric collection settings. * change description of validate OTE * fix a typo in the command name and add a delay before we try the unmount * Use underscore as separator for flags in place of hyphens * Enable host_metrics and disable reliability_metrics by default in configure OTE * Collect reliability metrics in the free namespace * Remove user from cmd params for HANA Replication * Enable workload manager metric collection by default. * Add support configuration flag to enable legacy WLM metric data submission workflow. * Lowers the log level of discovery to info * Fix for HANA Replication Config * Add additional instance-id parameter for users who do not want to provide port number * Use _ instead of - for parameters in configurebackint * Implementing panic recovery to HANA Monitoring: CreateWorkerPool * Fix issue with process metrics subroutine starting. * Add a flag to enable or disable workload discovery. * Reduce logs in sapdiscovery to debug, these are now run a lot more frequently and are flooding the logs * Use bucket `cloudsapdeploystaging` for staging environment. * Updates default value handling for system discovery flag. * Added default values to some frequency flags in configure OTE * force a sync before unmounting to clear out stale file handles * Retain recoverable routine in process metrics. * Ensures slow metrics workers stop on context cancellation.
John Paul Adrian Glaubitz (glaubitz)
accepted
request 1141503
from
John Paul Adrian Glaubitz (glaubitz)
(revision 9)
- Update to version 3.0 (bsc#1218736, bsc#1218737) * Suppress packemaker command error to debug to avoid log flooding * Expand load balancing cluster discovery. * Log success messages in OTEs to STDOUT instead of STDERR used by log.Print * Use bash always to avoid variation of behavior across OS/Shell types * Minor updates to installbackint. * Backint compose step properly saves metadata. * Fix issue with discovery on ASCS instances. * hanadiskrestore - fix the format of disktype string for disk create API * Fix issue with PCS cluster address discovery. * Update transform to insight * Rename HANA backup/restore OTEs to reflect they are supported for all disks and not just persistent disk * Increase the timeout for HDB stop to account for busy DBs * Adding project sap-ecs-testing to the list. * PD Restore - Support provisioned-iops and provisioned-throughput * Integration test for configure OTE * Added precondition in hana pd backup for stripped LVM * Add a precondition check to verify user has passed a valid snapshot name that is present in the current project * Update the usage to reflect additional required param * Minor path update for supportbundle OTE. * Fixing bug in slow moving metrics partial collection scenarios * Adding check for agent status after restart. * Ensure Backint ComposeChunks has a valid bucket handle * Discover whether a Netweaver instance is ABAP or Java * Replace standard slices package with third party version * WLM HANA metric `ha_in_same_zone` now reports instance names for HA nodes in the same zone * Fix data race condition for Backint Backup with new client connections
Robert Schweikert (rjschwei)
accepted
request 1129481
from
John Paul Adrian Glaubitz (glaubitz)
(revision 8)
- Update to version 2.8 (bsc#1217373, bsc#1217374) * Bump agent version to 2.8 to support C3/M3 certification * Update go.yml to use go 1.21 * Switch from "slices" to "go_exp.../slices" for go version dependency * Use newly refactored discovery packages. * Fixes issue with diskname from source or device name * Adds extreme disk type IOps and Throughput for host metrics * Add `INTEGRATION` target config environment for collection definition testing * Add project number to SAP System proto * Add a cache to discovered resources. This reduces the number of API calls needed to perform System Discovery * Replace windows wmic hardware queries with PowerShell wmi queries * Fix test flakiness * Improve development process for collection definition configuration * HANA PD based snapshot and restore - changes to add wait for uploading * Fix for kokoro build issue in processmetrics/networkstats * GCBDR SAPCoreAPP Package in Agent for SAP * Add version tracking for WLM validation config * Send workload validation config to remote instances for use during remote collection * Add flag for passing in workload validation config into remote collection OTE * Bump google.golang.org/grpc from 1.58.2 to 1.58.3 - from version 2.7 * Added ote for hma dashboards migration * Increase Max backoff in storage package to 300 seconds * Added subpaths for collection of required TCP metrics * Add more debug logs and increase the wait-time for PD operations in restore * No public description * Add 30 second timeout to read/write from the local file system for Backint * No public description * Adds RHEL 9 VM Manager policy
Robert Schweikert (rjschwei)
accepted
request 1119205
from
John Paul Adrian Glaubitz (glaubitz)
(revision 7)
- Update to version 2.6 (bsc#1215672, bsc#1215673) * Rolling back previous change for storing Project Number, Project ID is sufficient, no need to add complexity * Determine location of HANA global.ini using SAP system discovery logic * Add numeric project ID prefix to object name for ReadMetrics * Discovery now looks up and stores project number with discovery data * ReadMetrics updates for IAM permissions and bucket object names * fixing the bug in backoff logic, using separate policies for each collector and adding some logs * Backint migration from the old agent and supporting legacy parameters * adding new backoff policies for process metrics and fixing the bug in process metrics sapservice collector * Bump SAP Agent version to 2.6 * Fix an issue where HANA hosts may not be discovered properly if hostname differs from instance name * Use Go 1.20 friendly sorting solution * adding retries in process metrics logic with backoffs * Fix parsing of instance (host/VM) name in Pacemaker pcmk_delay_max metric * Add the collection definition changes for the SAP HANA Topology metrics * Template for Cloud Monitoring Alerts for Backint errors * adding backoff to InstanceProperties to each collector * Reduced the number of parameters of startXX functions by consolidating them into respective structs * completing TODO (b/298315981): Create a map from skipped list metrics and pass it to collectors. * Proto package name changes to reflect the current path * Use instance_name instead of instance_id for baremetal systems * Decode encryption keys for Backint. * Moving hareplication metric to fast moving metrics * Added backoffs package in process metrics to keep the backoff policies
Robert Schweikert (rjschwei)
accepted
request 1108231
from
John Paul Adrian Glaubitz (glaubitz)
(revision 6)
- Update to version 2.5 + No upstream changelog provided
Robert Schweikert (rjschwei)
accepted
request 1104345
from
John Paul Adrian Glaubitz (glaubitz)
(revision 5)
- Install missing default configuration file (bsc#1214293)
Robert Schweikert (rjschwei)
accepted
request 1099447
from
John Paul Adrian Glaubitz (glaubitz)
(revision 4)
- Update to version 2.1 (bsc#1213397) + No upstream changelog provided
Sean Marlow (seanmarlow)
accepted
request 1098311
from
John Paul Adrian Glaubitz (glaubitz)
(revision 3)
- Update to version 2.0 (bsc#1211516, bsc#1211517) + No upstream changelog provided
John Paul Adrian Glaubitz (glaubitz)
accepted
request 1098308
from
John Paul Adrian Glaubitz (glaubitz)
(revision 2)
- Update to version 1.5.1 (bsc#1210464, bsc#1210465) + No upstream changelog provided - Raise golang API version to 1.20
John Paul Adrian Glaubitz (glaubitz)
accepted
request 1098307
from
John Paul Adrian Glaubitz (glaubitz)
(revision 1)
- Initial build (#jsc-PED-2879) + Version 1.0
Displaying all 15 revisions