Unreduce max num instances for batch parsing

Increase max instances (#964)

Reduce legacy parser instance counts

Add NDT5ResultRowStandardColumns (#956) …
Site annotator for synthetic UUID (#958) …
fix bug after go package update (#960) …
reduce to 40 instances to reduce legacy cost (#959) …
reduce to 4-6 instances, smaller memory (#962) …
Remove client and server annotations from ndt5 schema (#963)

Adds 10s to all DISCOv1 sample timestamps

This change will make DISCOv1 timestamps represent the same thing as DISCOv2 timestamps: then timestamp of the end of a 10s sample window, not the beginning of the window.

Modifies SwitchStats schema + removes Cloud Functions

u-prod-v2.5.3...prod-v2.5.1

DISCOv2 is introducing a new field into each sample, counter, to record raw SNMP counter values (instead of just the diff from the previous value). This release modifies to the switch data type schema to include this new field.

Additionally, this release modifies update.go so that the switch schema can be modified automatically by scripts, instead of someone having to manually modify tables.

Finally, this release removes all Cloud Functions pieces from the Travis configuration, as they were no longer used.

Fix prometheus monitoring

This correctly initializes the prometheus monitoring port on :9090

Add GCS metadata & use milliseconds for ndt7 MinRTT

Add standard and legacy update-schema types (#935) …
Migrate to go/cloud/gcs (#936) …
Add row count and error meta-data to json output files (#937) …
Use milliseconds not seconds (#938)

Release prometheus fix to legacy batch processor

fix typo (#942)

* fix typo

Universal parser with table schema updates.

This adds the travis deployment rules to update or create the ndt7 and annotation tables.

Production universal parser & restore standard queue_pusher

Since last prod deploy:

Fix labels, add commit and release tags (#878) …
Add JSON parsing for new traceroute output format (#877) …
Implement archive filter (#879) …
Deploy sandbox k8s parsers only on u-sandbox branches (#885) …
retire ALL daily pipeline (#886) …
Restore ndt5 schema updates (#890)
fix bad deployment condition (#891) …
Add ID field to ndt7 result row schema (#882) …
Update NDT7 and Annotation schemas to use Standard Columns (#895) …
Use GCS archive path date for row.Date (#896) …
Add sink.close (#898) …
Build and ship in same environment (#903)
Metrics and misc (#901) …
Update ndt5 schema (#904) …
Add Fail fast option in ProcessAll, so that file write errors can be … …
Add pathAndFilename, and NewSinkFactory (#910) …
update ndt7 result row schema (#907) …
Add some missing BBR variable descriptions, misc fixes and updated bi… …
Schema docs update2 (#916) …
refactor main, unit test, more status info (#911) …
Revert "refactor main, unit test, more status info (#911)" (#917) …
add ctx to Run() (#919) …
conditional json output and enhanced status (#920) …
Accepts JSON and JSONL suffixes for test filenames. (#922)
Refactor main, add unit test (#921) …
Enable GCS json output (#923) …
Uses whole sample for DISCOv2. (#924) …
k8s config to use 7 cpus, to allow mapping to "parser-pool" (#925) …
Remove testdata files with invalid characters in filename (#926) …
Restore deploy and migrate queue_pusher to go111 (#928) …
Add prod deployment rule (#930) …

Update unittests

Refactor main, unit test, more status info (#911) …
Revert "refactor main, unit test, more status info (#911)" (#917) …
add ctx to Run() (#919) …
conditional json output and enhanced status (#920) …
Accepts JSON and JSONL suffixes for test filenames. (#922)
Refactor main, add unit test (#921) …
Enable GCS json output (#923) …
Uses whole sample for DISCOv2. (#924) …
k8s config to use 7 cpus, to allow mapping to "parser-pool" (#925) …
Remove testdata files with invalid characters in filename (#926) …