Skip to content
Projects
Groups
Snippets
Help
Loading...
Help
Support
Keyboard shortcuts
?
Submit feedback
Contribute to GitLab
Sign in / Register
Toggle navigation
N
neoppod
Project overview
Project overview
Details
Activity
Releases
Repository
Repository
Files
Commits
Branches
Tags
Contributors
Graph
Compare
Issues
0
Issues
0
List
Boards
Labels
Milestones
Merge Requests
0
Merge Requests
0
CI / CD
CI / CD
Pipelines
Jobs
Schedules
Analytics
Analytics
CI / CD
Repository
Value Stream
Wiki
Wiki
Snippets
Snippets
Members
Members
Collapse sidebar
Close sidebar
Activity
Graph
Create a new issue
Jobs
Commits
Issue Boards
Open sidebar
Levin Zimmermann
neoppod
Commits
083aed66
Commit
083aed66
authored
Dec 17, 2018
by
Kirill Smelkov
Browse files
Options
Browse Files
Download
Email Patches
Plain Diff
.
parent
9003ac50
Changes
1
Hide whitespace changes
Inline
Side-by-side
Showing
1 changed file
with
91 additions
and
81 deletions
+91
-81
go/zodb/storage/fs1/filestorage.go
go/zodb/storage/fs1/filestorage.go
+91
-81
No files found.
go/zodb/storage/fs1/filestorage.go
View file @
083aed66
...
@@ -101,7 +101,7 @@ type FileStorage struct {
...
@@ -101,7 +101,7 @@ type FileStorage struct {
txnhMax
TxnHeader
txnhMax
TxnHeader
// driver client <- watcher: data file updates.
// driver client <- watcher: data file updates.
watchq
chan
interface
{}
// XXX
watchq
chan
watchEvent
}
}
// IStorageDriver
// IStorageDriver
...
@@ -450,14 +450,14 @@ func (fs *FileStorage) watcher() {
...
@@ -450,14 +450,14 @@ func (fs *FileStorage) watcher() {
_
=
err
_
=
err
}
}
// watch watches updates to
fs.file and notifies subscriber about new transaction via fs.watchq
.
// watch watches updates to
.file and notifies Watch about new transactions
.
//
//
// watch is the only place that mutates index and txnh{Min,Max}.
// watch is the only place that mutates index and txnh{Min,Max}.
// XXX ^^^ will change after commit is implemented.
// XXX ^^^ will change after commit is implemented.
func
(
fs
*
FileStorage
)
watch
()
(
err
error
)
{
func
(
fs
*
FileStorage
)
watch
()
(
err
error
)
{
f
:=
fs
.
file
f
:=
fs
.
file
idx
:=
fs
.
index
idx
:=
fs
.
index
defer
xerr
.
Contextf
(
&
err
,
"%s: watch"
,
f
.
Name
())
defer
xerr
.
Contextf
(
&
err
,
"%s: watch
er
"
,
f
.
Name
())
// XXX race -> start watching before FileStorage is returned to user
// XXX race -> start watching before FileStorage is returned to user
...
@@ -477,15 +477,10 @@ func (fs *FileStorage) watch() (err error) {
...
@@ -477,15 +477,10 @@ func (fs *FileStorage) watch() (err error) {
//
//
// besides relying on notify we also check file periodically to avoid
// besides relying on notify we also check file periodically to avoid
// stalls due to e.g. OS notification errors.
// stalls due to e.g. OS notification errors.
var
it
*
Iter
itReset
:=
func
()
{
// reset it to (re)start from toppos with empty buffer
it
=
Iterate
(
seqReadAt
(
f
),
idx
.
TopPos
,
IterForward
)
}
itReset
()
tick
:=
time
.
NewTicker
(
1
*
time
.
Second
)
tick
:=
time
.
NewTicker
(
1
*
time
.
Second
)
defer
tick
.
Stop
()
defer
tick
.
Stop
()
var
t0partial
time
.
Time
var
t0partial
time
.
Time
mainloop
:
for
{
for
{
// XXX select here
// XXX select here
...
@@ -503,90 +498,91 @@ func (fs *FileStorage) watch() (err error) {
...
@@ -503,90 +498,91 @@ func (fs *FileStorage) watch() (err error) {
return
fmt
.
Errorf
(
"file truncated (%d -> %d)"
,
idx
.
TopPos
,
fsize
)
return
fmt
.
Errorf
(
"file truncated (%d -> %d)"
,
idx
.
TopPos
,
fsize
)
}
}
// there is some data after toppos - try to read it.
// there is some data after toppos - try to advance as much as we can.
err
=
it
.
NextTxn
(
LoadNoStrings
)
// start iterating afresh with empty buffer.
if
err
!=
nil
{
it
:=
Iterate
(
seqReadAt
(
f
),
idx
.
TopPos
,
IterForward
)
// transaction header could not be fully read.
for
{
//
err
=
it
.
NextTxn
(
LoadNoStrings
)
// Even though FileStorage code always calls write with full txn
if
err
!=
nil
{
// header, the kernel could do the write in parts, e.g. if written
// transaction header could not be fully read.
// region overlaps page boundary.
//
//
// Even though FileStorage code always calls write with full txn
// we check for some time to distinguish in-progress write from just
// header, the kernel could do the write in parts, e.g. if written
// trailing garbage.
// region overlaps page boundary.
if
errors
.
Cause
(
err
)
==
io
.
ErrUnexpectedEOF
{
//
now
:=
time
.
Now
()
// we check for some time to distinguish in-progress write from just
if
t0partial
.
IsZero
()
{
// trailing garbage.
t0partial
=
now
if
errors
.
Cause
(
err
)
==
io
.
ErrUnexpectedEOF
{
}
else
if
now
.
Sub
(
t0partial
)
>
3
*
time
.
Second
{
now
:=
time
.
Now
()
return
err
// garbage
if
t0partial
.
IsZero
()
{
}
t0partial
=
now
}
else
{
}
else
if
now
.
Sub
(
t0partial
)
>
3
*
time
.
Second
{
// not ok - e.g. IO error
return
err
// garbage
// only EOF is ok - it is e.g. when transaction was aborted
}
if
err
!=
io
.
EOF
{
}
else
{
return
err
// not ok - e.g. IO error
// only EOF is ok - it can happen when transaction was aborted
if
err
!=
io
.
EOF
{
return
err
}
// EOF - reset t₀(partial)
t0partial
=
time
.
Time
{}
}
}
// EOF - reset t₀(partial)
// after any error (EOF, partial read) we have to resync
t0partial
=
time
.
Time
{}
continue
mainloop
}
}
// after any error (EOF, partial read) we have to retry with
// read ok - reset t₀(partial)
// reset bufferring.
t0partial
=
time
.
Time
{}
itReset
()
continue
}
// read ok - reset t₀(partial)
t0partial
=
time
.
Time
{}
// XXX dup wrt Index.Update
// XXX dup wrt Index.Update
// we could successfully read the transaction header. Try to see now,
// we could successfully read the transaction header. Try to see now,
// whether it is finished transaction or not.
// whether it is finished transaction or not.
if
it
.
Txnh
.
Status
==
zodb
.
TxnInprogress
{
if
it
.
Txnh
.
Status
==
zodb
.
TxnInprogress
{
// not yet. we have to reset because transaction finish writes
// not yet. we have to resync because transaction finish writes
// to what we have already buffered.
// to what we have already buffered.
itReset
()
continue
mainloop
continue
}
}
// it is fully-committed transaction. scan its data records to update
// it is fully-committed transaction. scan its data records to update
// our index & notify watchers. There is no expected errors here.
// our index & notify watchers. There is no expected errors here.
var
oidv
[]
zodb
.
Oid
var
oidv
[]
zodb
.
Oid
update
:=
map
[
zodb
.
Oid
]
int64
{}
update
:=
map
[
zodb
.
Oid
]
int64
{}
for
{
for
{
err
=
it
.
NextData
()
err
=
it
.
NextData
()
if
err
!=
nil
{
err
=
okEOF
(
err
)
if
err
!=
nil
{
if
err
!=
nil
{
return
err
err
=
okEOF
(
err
)
if
err
!=
nil
{
return
err
}
break
}
}
break
update
[
it
.
Datah
.
Oid
]
=
it
.
Datah
.
Pos
oidv
=
append
(
oidv
,
it
.
Datah
.
Oid
)
}
}
update
[
it
.
Datah
.
Oid
]
=
it
.
Datah
.
Pos
// update index & txnh{MinMax}
oidv
=
append
(
oidv
,
it
.
Datah
.
Oid
)
fs
.
mu
.
Lock
()
}
idx
.
TopPos
=
it
.
Txnh
.
Pos
+
it
.
Txnh
.
Len
for
oid
,
pos
:=
range
update
{
idx
.
Set
(
oid
,
pos
)
}
fs
.
txnhMax
.
CloneFrom
(
&
it
.
Txnh
)
if
fs
.
txnhMin
.
Len
==
0
{
// was empty
fs
.
txnhMin
.
CloneFrom
(
&
it
.
Txnh
)
}
fs
.
mu
.
Unlock
()
// update index & txnh{MinMax}
// XXX cancel on close
fs
.
mu
.
Lock
()
fs
.
watchq
<-
watchEvent
{
it
.
Txnh
.
Tid
,
oidv
}
idx
.
TopPos
=
it
.
Txnh
.
Pos
+
it
.
Txnh
.
Len
for
oid
,
pos
:=
range
update
{
idx
.
Set
(
oid
,
pos
)
}
fs
.
txnhMax
.
CloneFrom
(
&
it
.
Txnh
)
if
fs
.
txnhMin
.
Len
==
0
{
// was empty
fs
.
txnhMin
.
CloneFrom
(
&
it
.
Txnh
)
}
}
fs
.
mu
.
Unlock
()
// XXX -> watchq
select
{
select
{
// XXX handle
quit
// XXX handle
close
case
err
:=
<-
w
.
Errors
:
case
err
:=
<-
w
.
Errors
:
if
err
==
fsnotify
.
ErrEventOverflow
{
if
err
==
fsnotify
.
ErrEventOverflow
{
...
@@ -609,9 +605,23 @@ func (fs *FileStorage) watch() (err error) {
...
@@ -609,9 +605,23 @@ func (fs *FileStorage) watch() (err error) {
}
}
}
}
func
(
fs
*
FileStorage
)
Watch
(
ctx
context
.
Context
)
(
zodb
.
Tid
,
[]
zodb
.
Oid
,
error
)
{
// watchEvent is one event from watch to Watch
panic
(
"TODO"
)
type
watchEvent
struct
{
// <-fs.watchq
tid
zodb
.
Tid
oidv
[]
zodb
.
Oid
}
func
(
fs
*
FileStorage
)
Watch
(
ctx
context
.
Context
)
(
_
zodb
.
Tid
,
_
[]
zodb
.
Oid
,
err
error
)
{
defer
xerr
.
Contextf
(
&
err
,
"%s: watch"
,
fs
.
file
.
Name
())
// XXX handle close
select
{
case
<-
ctx
.
Done
()
:
return
zodb
.
InvalidTid
,
nil
,
ctx
.
Err
()
case
w
:=
<-
fs
.
watchq
:
return
w
.
tid
,
w
.
oidv
,
nil
}
}
}
// --- open + rebuild index ---
// --- open + rebuild index ---
...
...
Write
Preview
Markdown
is supported
0%
Try again
or
attach a new file
Attach a file
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Cancel
Please
register
or
sign in
to comment