rust/neon
1
0
Fork 0
mirror of https://github.com/neondatabase/neon.git synced 2026-05-14 11:40:38 +00:00
Code Issues Packages Projects Releases Wiki Activity

Print more information about context of failed walredo requests (#3003)

Browse Source
This commit is contained in:
Konstantin Knizhnik
2022-12-08 09:12:38 +02:00
committed by GitHub
parent b50e0793cf
commit e1ef62f086
4 changed files with 16 additions and 10 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
pageserver/benches/bench_walredo.rs
View File

@@ -431,7 +431,7 @@ fn pg_record(will_init: bool, bytes: &'static [u8]) -> NeonWalRecord {
struct Request {
key: Key,
lsn: Lsn,
base_img: Option<Bytes>,
base_img: Option<(Lsn, Bytes)>,
records: Vec<(Lsn, NeonWalRecord)>,
pg_version: u32,
}

2
pageserver/src/tenant.rs
View File

@@ -2669,7 +2669,7 @@ pub mod harness {
&self,
key: Key,
lsn: Lsn,
base_img: Option<Bytes>,
base_img: Option<(Lsn, Bytes)>,
records: Vec<(Lsn, NeonWalRecord)>,
_pg_version: u32,
) -> Result<Bytes, WalRedoError> {

6
pageserver/src/tenant/timeline.rs
View File

@@ -2642,24 +2642,22 @@ impl Timeline {
data.records.len()
);
} else {
let base_img = if let Some((_lsn, img)) = data.img {
if data.img.is_some() {
trace!(
"found {} WAL records and a base image for {} at {}, performing WAL redo",
data.records.len(),
key,
request_lsn
);
Some(img)
} else {
trace!("found {} WAL records that will init the page for {} at {}, performing WAL redo", data.records.len(), key, request_lsn);
None
};
let last_rec_lsn = data.records.last().unwrap().0;
let img = self
.walredo_mgr
.request_redo(key, request_lsn, base_img, data.records, self.pg_version)
.request_redo(key, request_lsn, data.img, data.records, self.pg_version)
.context("Failed to reconstruct a page image:")?;
if img.len() == page_cache::PAGE_SZ {

16
pageserver/src/walredo.rs
View File

@@ -84,7 +84,7 @@ pub trait WalRedoManager: Send + Sync {
&self,
key: Key,
lsn: Lsn,
base_img: Option<Bytes>,
base_img: Option<(Lsn, Bytes)>,
records: Vec<(Lsn, NeonWalRecord)>,
pg_version: u32,
) -> Result<Bytes, WalRedoError>;
@@ -147,7 +147,7 @@ impl WalRedoManager for PostgresRedoManager {
&self,
key: Key,
lsn: Lsn,
base_img: Option<Bytes>,
base_img: Option<(Lsn, Bytes)>,
records: Vec<(Lsn, NeonWalRecord)>,
pg_version: u32,
) -> Result<Bytes, WalRedoError> {
@@ -156,7 +156,8 @@ impl WalRedoManager for PostgresRedoManager {
return Err(WalRedoError::InvalidRequest);
}
let mut img: Option<Bytes> = base_img;
let base_img_lsn = base_img.as_ref().map(|p| p.0).unwrap_or(Lsn::INVALID);
let mut img = base_img.map(|p| p.1);
let mut batch_neon = can_apply_in_neon(&records[0].1);
let mut batch_start = 0;
for i in 1..records.len() {
@@ -170,6 +171,7 @@ impl WalRedoManager for PostgresRedoManager {
key,
lsn,
img,
base_img_lsn,
&records[batch_start..i],
self.conf.wal_redo_timeout,
pg_version,
@@ -189,6 +191,7 @@ impl WalRedoManager for PostgresRedoManager {
key,
lsn,
img,
base_img_lsn,
&records[batch_start..],
self.conf.wal_redo_timeout,
pg_version,
@@ -223,11 +226,13 @@ impl PostgresRedoManager {
///
/// Process one request for WAL redo using wal-redo postgres
///
#[allow(clippy::too_many_arguments)]
fn apply_batch_postgres(
&self,
key: Key,
lsn: Lsn,
base_img: Option<Bytes>,
base_img_lsn: Lsn,
records: &[(Lsn, NeonWalRecord)],
wal_redo_timeout: Duration,
pg_version: u32,
@@ -282,9 +287,12 @@ impl PostgresRedoManager {
// next request will launch a new one.
if result.is_err() {
error!(
"error applying {} WAL records ({} bytes) to reconstruct page image at LSN {}",
"error applying {} WAL records {}..{} ({} bytes) to base image with LSN {} to reconstruct page image at LSN {}",
records.len(),
records.first().map(|p| p.0).unwrap_or(Lsn(0)),
records.last().map(|p| p.0).unwrap_or(Lsn(0)),
nbytes,
base_img_lsn,
lsn
);
let process = process_guard.take().unwrap();