fix wrong regex of allowed errors in test_ondemand_download_failure_to_replace

Signed-off-by: Alex Chi Z <chi@neon.tech>

control_plane/src/endpoint.rs

@@ -564,9 +564,7 @@ impl Endpoint {
                 }
                 Err(e) => {
                     if attempt == MAX_ATTEMPTS {
-                        return Err(e).context(
-                            "timed out waiting to connect to compute_ctl HTTP; last error: {e}",
-                        );
+                        return Err(e).context("timed out waiting to connect to compute_ctl HTTP");
                     }
                 }
             }

pageserver/src/tenant.rs

@@ -46,7 +46,6 @@ use std::sync::{Mutex, RwLock};
 use std::time::{Duration, Instant};
 
 use self::config::TenantConf;
-use self::delete::DeleteTimelineFlow;
 use self::metadata::LoadMetadataError;
 use self::metadata::TimelineMetadata;
 use self::remote_timeline_client::RemoteTimelineClient;
@@ -70,6 +69,7 @@ use crate::tenant::storage_layer::ImageLayer;
 use crate::tenant::storage_layer::Layer;
 use crate::InitializationOrder;
 
+use crate::tenant::timeline::delete::DeleteTimelineFlow;
 use crate::tenant::timeline::uninit::cleanup_timeline_directory;
 use crate::virtual_file::VirtualFile;
 use crate::walredo::PostgresRedoManager;
@@ -117,7 +117,6 @@ mod remote_timeline_client;
 pub mod storage_layer;
 
 pub mod config;
-pub mod delete;
 pub mod mgr;
 pub mod tasks;
 pub mod upload_queue;

pageserver/src/tenant/mgr.rs

@@ -26,7 +26,7 @@ use crate::{InitializationOrder, IGNORED_TENANT_FILE_NAME};
 use utils::fs_ext::PathExt;
 use utils::id::{TenantId, TimelineId};
 
-use super::delete::DeleteTimelineFlow;
+use super::timeline::delete::DeleteTimelineFlow;
 
 /// The tenants known to the pageserver.
 /// The enum variants are used to distinguish the different states that the pageserver can be in.

pageserver/src/tenant/timeline.rs

@@ -1,3 +1,4 @@
+pub mod delete;
 mod eviction_task;
 pub mod layer_manager;
 mod logical_size;
@@ -79,6 +80,7 @@ use crate::METADATA_FILE_NAME;
 use crate::ZERO_PAGE;
 use crate::{is_temporary, task_mgr};
 
+use self::delete::DeleteTimelineFlow;
 pub(super) use self::eviction_task::EvictionTaskTenantState;
 use self::eviction_task::EvictionTaskTimelineState;
 use self::layer_manager::LayerManager;
@@ -86,7 +88,6 @@ use self::logical_size::LogicalSize;
 use self::walreceiver::{WalReceiver, WalReceiverConf};
 
 use super::config::TenantConf;
-use super::delete::DeleteTimelineFlow;
 use super::remote_timeline_client::index::IndexPart;
 use super::remote_timeline_client::RemoteTimelineClient;
 use super::storage_layer::{

pageserver/src/tenant/timeline/delete.rs

@@ -15,15 +15,17 @@ use utils::{
 use crate::{
     config::PageServerConf,
     task_mgr::{self, TaskKind},
-    tenant::{remote_timeline_client, DeleteTimelineError},
+    tenant::{
+        metadata::TimelineMetadata,
+        remote_timeline_client::{
+            self, PersistIndexPartWithDeletedFlagError, RemoteTimelineClient,
+        },
+        CreateTimelineCause, DeleteTimelineError, Tenant,
+    },
     InitializationOrder,
 };
 
-use super::{
-    metadata::TimelineMetadata,
-    remote_timeline_client::{PersistIndexPartWithDeletedFlagError, RemoteTimelineClient},
-    CreateTimelineCause, Tenant, Timeline,
-};
+use super::Timeline;
 
 /// Now that the Timeline is in Stopping state, request all the related tasks to shut down.
 async fn stop_tasks(timeline: &Timeline) -> Result<(), DeleteTimelineError> {

test_runner/fixtures/neon_fixtures.py

@@ -1504,6 +1504,7 @@ class NeonCli(AbstractNeonCli):
         safekeepers: Optional[List[int]] = None,
         tenant_id: Optional[TenantId] = None,
         lsn: Optional[Lsn] = None,
+        branch_name: Optional[str] = None,
     ) -> "subprocess.CompletedProcess[str]":
         args = [
             "endpoint",
@@ -1517,8 +1518,11 @@ class NeonCli(AbstractNeonCli):
             args.append(f"--lsn={lsn}")
         args.extend(["--pg-port", str(pg_port)])
         args.extend(["--http-port", str(http_port)])
+
         if safekeepers is not None:
             args.extend(["--safekeepers", (",".join(map(str, safekeepers)))])
+        if branch_name is not None:
+            args.extend(["--branch-name", branch_name])
         if endpoint_id is not None:
             args.append(endpoint_id)
 

test_runner/regress/test_neon_local_cli.py

@@ -16,11 +16,13 @@ def test_neon_cli_basics(neon_env_builder: NeonEnvBuilder, port_distributor: Por
             endpoint_id="ep-basic-main", pg_port=pg_port, http_port=http_port
         )
 
-        env.neon_cli.create_branch(new_branch_name="migration_check")
+        branch_name = "migration-check"
+
+        env.neon_cli.create_branch(new_branch_name=branch_name)
         pg_port = port_distributor.get_port()
         http_port = port_distributor.get_port()
         env.neon_cli.endpoint_start(
-            endpoint_id="ep-migration_check", pg_port=pg_port, http_port=http_port
+            f"ep-{branch_name}", pg_port, http_port, branch_name=branch_name
         )
     finally:
         env.neon_cli.stop()

test_runner/regress/test_ondemand_download.py

@@ -712,10 +712,10 @@ def test_ondemand_download_failure_to_replace(
     env.pageserver.allowed_errors.append(actual_message)
 
     env.pageserver.allowed_errors.append(
-        ".* ERROR .*Error processing HTTP request: InternalServerError\\(get local timeline info"
+        ".* ERROR .*Error processing HTTP request: InternalServerError\\(get local timeline info.*"
     )
     # this might get to run and attempt on-demand, but not always
-    env.pageserver.allowed_errors.append(".* ERROR .*Task 'initial size calculation'")
+    env.pageserver.allowed_errors.append(".* ERROR .*Task 'initial size calculation'.*")
 
     # if the above returned, then we didn't have a livelock, and all is well
 

test_runner/regress/test_remote_storage.py

@@ -348,6 +348,9 @@ def test_remote_storage_upload_queue_retries(
     # XXX: should vary this test to selectively fail just layer uploads, index uploads, deletions
     #      but how do we validate the result after restore?
 
+    # these are always possible when we do an immediate stop. perhaps something with compacting has changed since.
+    env.pageserver.allowed_errors.append(r".*found future (delta|image) layer.*")
+
     env.pageserver.stop(immediate=True)
     env.endpoints.stop_all()