Skip to content

[fix](regression) Stabilize delete bitmap cache metrics test #65535

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.

Sign up for GitHub

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

Draft
shuke987 wants to merge 2 commits into
base: master
Choose a base branch
from
Draft

[fix](regression) Stabilize delete bitmap cache metrics test #65535

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
bfde4a3
[fix](regression) Stabilize delete bitmap cache metrics test
shuke987 Jul 13, 2026
10ef490
[fix](regression) Pin cloud delete bitmap cache checks
shuke987 Jul 17, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
157 changes: 91 additions & 66 deletions regression-test/suites/metrics_p0/test_delete_bitmap_metrics.groovy
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -17,7 +17,7 @@

import org.codehaus.groovy.runtime.IOGroovyMethods

suite("test_delete_bitmap_metrics", "p0") {
suite("test_delete_bitmap_metrics", "p0,nonConcurrent") {

@github-actions github-actions Bot Jul 17, 2026

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Adding nonConcurrent makes this suite a SINGLE suite, but regression-test/pipeline/cloud_p0/run.sh invokes the framework with -runNonConcurrent false, so RegressionTest.runScripts returns before this suite can run. The standalone nonConcurrent pipeline is shared-nothing, and vault P0 only scans vault_p0, so the new isCloudMode()/multi-replica fix loses its checked-in cloud coverage. Please keep the global config mutation serialized while retaining a cloud execution path, for example with a cloud SINGLE phase/job or a deterministic binding that does not require reclassifying this suite.

def backendId_to_backendIP = [:]
def backendId_to_backendHttpPort = [:]
def backendId_to_params = [string: [:]]
Expand Down Expand Up @@ -183,80 +183,105 @@ suite("test_delete_bitmap_metrics", "p0") {
qt_sql "select * from ${testTable} order by plan_id"


def tablets = sql_return_maparray """ show tablets from ${testTable}; """
logger.info("tablets: " + tablets)
def local_delete_bitmap_count = 0
def ms_delete_bitmap_count = 0
def local_delete_bitmap_cardinality = 0;
def ms_delete_bitmap_cardinality = 0;
for (def tablet in tablets) {
String tablet_id = tablet.TabletId
def tablet_info = sql_return_maparray """ show tablet ${tablet_id}; """
logger.info("tablet: " + tablet_info)
String trigger_backend_id = tablet.BackendId
def checkPerReplicaDeleteBitmap = {
def tablets = sql_return_maparray """ show tablets from ${testTable}; """
tablets.sort { lhs, rhs -> lhs.ReplicaId.toLong() <=> rhs.ReplicaId.toLong() }
logger.info("tablets: " + tablets)
def local_delete_bitmap_count = 0
def ms_delete_bitmap_count = 0
def local_delete_bitmap_cardinality = 0;
def ms_delete_bitmap_cardinality = 0;
try {
for (int ri = 0; ri < tablets.size(); ri++) {
// use_fix_replica selects replicas in ReplicaId order. Populate and
// inspect each replica immediately because the agg cache is an LRU.
sql "set use_fix_replica=${ri};"
sql "select * from ${testTable};"

// before compaction, delete_bitmap_count is (rowsets num - 1)
local_delete_bitmap_count = getLocalDeleteBitmapStatus(backendId_to_backendIP[trigger_backend_id], backendId_to_backendHttpPort[trigger_backend_id], tablet_id).delete_bitmap_count
local_delete_bitmap_cardinality = getLocalDeleteBitmapStatus(backendId_to_backendIP[trigger_backend_id], backendId_to_backendHttpPort[trigger_backend_id], tablet_id).cardinality
logger.info("local_delete_bitmap_count:" + local_delete_bitmap_count)
logger.info("local_delete_bitmap_cardinality:" + local_delete_bitmap_cardinality)
assertTrue(local_delete_bitmap_count == 7)
assertTrue(local_delete_bitmap_cardinality == 7)
def tablet = tablets[ri]
String tablet_id = tablet.TabletId
def tablet_info = sql_return_maparray """ show tablet ${tablet_id}; """
logger.info("tablet: " + tablet_info)
String trigger_backend_id = tablet.BackendId

if (isCloudMode()) {
ms_delete_bitmap_count = getMSDeleteBitmapStatus(backendId_to_backendIP[trigger_backend_id], backendId_to_backendHttpPort[trigger_backend_id], tablet_id).delete_bitmap_count
ms_delete_bitmap_cardinality = getMSDeleteBitmapStatus(backendId_to_backendIP[trigger_backend_id], backendId_to_backendHttpPort[trigger_backend_id], tablet_id).cardinality
logger.info("ms_delete_bitmap_count:" + ms_delete_bitmap_count)
logger.info("ms_delete_bitmap_cardinality:" + ms_delete_bitmap_cardinality)
assertTrue(ms_delete_bitmap_count == 7)
assertTrue(ms_delete_bitmap_cardinality == 7)
}
// before compaction, delete_bitmap_count is (rowsets num - 1)
local_delete_bitmap_count = getLocalDeleteBitmapStatus(backendId_to_backendIP[trigger_backend_id], backendId_to_backendHttpPort[trigger_backend_id], tablet_id).delete_bitmap_count
local_delete_bitmap_cardinality = getLocalDeleteBitmapStatus(backendId_to_backendIP[trigger_backend_id], backendId_to_backendHttpPort[trigger_backend_id], tablet_id).cardinality
logger.info("local_delete_bitmap_count:" + local_delete_bitmap_count)
logger.info("local_delete_bitmap_cardinality:" + local_delete_bitmap_cardinality)
assertTrue(local_delete_bitmap_count == 7)
assertTrue(local_delete_bitmap_cardinality == 7)

if (isCloudMode()) {
ms_delete_bitmap_count = getMSDeleteBitmapStatus(backendId_to_backendIP[trigger_backend_id], backendId_to_backendHttpPort[trigger_backend_id], tablet_id).delete_bitmap_count
ms_delete_bitmap_cardinality = getMSDeleteBitmapStatus(backendId_to_backendIP[trigger_backend_id], backendId_to_backendHttpPort[trigger_backend_id], tablet_id).cardinality
logger.info("ms_delete_bitmap_count:" + ms_delete_bitmap_count)
logger.info("ms_delete_bitmap_cardinality:" + ms_delete_bitmap_cardinality)
assertTrue(ms_delete_bitmap_count == 7)
assertTrue(ms_delete_bitmap_cardinality == 7)
}

def status = getAggCacheDeleteBitmapStatus(backendId_to_backendIP[trigger_backend_id], backendId_to_backendHttpPort[trigger_backend_id], tablet_id)
logger.info("agg cache status: ${status}")
assert status.delete_bitmap_count == 8
assert status.cardinality == 7
assert status.size > 0
def status = getAggCacheDeleteBitmapStatus(backendId_to_backendIP[trigger_backend_id], backendId_to_backendHttpPort[trigger_backend_id], tablet_id)
logger.info("agg cache status: ${status}")
assert status.delete_bitmap_count == 8
assert status.cardinality == 7
assert status.size > 0

status = getAggCacheDeleteBitmapStatus(backendId_to_backendIP[trigger_backend_id], backendId_to_backendHttpPort[trigger_backend_id], tablet_id, true)
logger.info("agg cache verbose status: ${status}")
status = getAggCacheDeleteBitmapStatus(backendId_to_backendIP[trigger_backend_id], backendId_to_backendHttpPort[trigger_backend_id], tablet_id, true)
logger.info("agg cache verbose status: ${status}")
}
} finally {
sql "set use_fix_replica=-1;"
}
}

def tablet_delete_bitmap_count = 0;
def base_rowset_delete_bitmap_count = 0;
int retry_time = 0;
while (retry_time < 10) {
log.info("retry_time: ${retry_time}")
getMetricsMethod.call() {
respCode, body ->
logger.info("test get delete bitmap count resp Code {}", "${respCode}".toString())
assertEquals("${respCode}".toString(), "200")
String out = "${body}".toString()
def strs = out.split('\n')
for (String line in strs) {
if (line.startsWith("tablet_max_delete_bitmap_score")) {
logger.info("find: {}", line)
tablet_delete_bitmap_count = line.replaceAll("tablet_max_delete_bitmap_score ", "").toInteger()
break
}
if (isCloudMode()) {
// Cloud multi-replica can resolve the same logical CloudReplica to a
// different physical BE for SHOW TABLETS and the warm-up scan. Disable
// it while binding the warm-up and per-BE cache assertion; the helper
// restores the original FE config even when an assertion fails.
setFeConfigTemporary([enable_cloud_multi_replica: false]) {
checkPerReplicaDeleteBitmap()
}
} else {
checkPerReplicaDeleteBitmap()
}

def tablet_delete_bitmap_count = 0;
def base_rowset_delete_bitmap_count = 0;
int retry_time = 0;
while (retry_time < 10) {
log.info("retry_time: ${retry_time}")
getMetricsMethod.call() {
respCode, body ->
logger.info("test get delete bitmap count resp Code {}", "${respCode}".toString())
assertEquals("${respCode}".toString(), "200")
String out = "${body}".toString()
def strs = out.split('\n')
for (String line in strs) {
if (line.startsWith("tablet_max_delete_bitmap_score")) {
logger.info("find: {}", line)
tablet_delete_bitmap_count = line.replaceAll("tablet_max_delete_bitmap_score ", "").toInteger()
break
}
for (String line in strs) {
if (line.startsWith("tablet_max_base_rowset_delete_bitmap_score")) {
logger.info("find: {}", line)
base_rowset_delete_bitmap_count = line.replaceAll("tablet_max_base_rowset_delete_bitmap_score ", "").toInteger()
break
}
}
for (String line in strs) {
if (line.startsWith("tablet_max_base_rowset_delete_bitmap_score")) {
logger.info("find: {}", line)
base_rowset_delete_bitmap_count = line.replaceAll("tablet_max_base_rowset_delete_bitmap_score ", "").toInteger()
break
}
}
if (tablet_delete_bitmap_count > 0 && base_rowset_delete_bitmap_count > 0) {
break;
} else {
Thread.sleep(10000)
retry_time++;
}
}
}
if (tablet_delete_bitmap_count > 0 && base_rowset_delete_bitmap_count > 0) {
break;
} else {
Thread.sleep(10000)
retry_time++;
}
assertTrue(tablet_delete_bitmap_count > 0)
assertTrue(base_rowset_delete_bitmap_count > 0)
}
assertTrue(tablet_delete_bitmap_count > 0)
assertTrue(base_rowset_delete_bitmap_count > 0)
} finally {
reset_be_param("check_tablet_delete_bitmap_interval_seconds")
}
Expand Down
Loading