TrueCloudLab
/
frostfs-node
18
1
Fork 20
Code Issues 88 Pull Requests 8 Actions Releases Activity

Refactor get service #193

Merged
fyrchik merged 7 commits from dstepanov-yadro/frostfs-node:refactoring/OBJECT-3610_getsvc into master 2023-04-05 14:38:49 +00:00
Conversation 1 Commits 7 Files Changed 14
dstepanov-yadro commented 2023-03-31 11:51:23 +00:00
Collaborator

Resolve linters for get service

Resolve linters for get service
dstepanov-yadro changed title from Refactor get service to WIP: Refactor get service 2023-03-31 11:51:32 +00:00
dstepanov-yadro force-pushed refactoring/OBJECT-3610_getsvc from d380229b0f to 5718390a84 2023-03-31 11:52:25 +00:00 Compare
dstepanov-yadro changed title from WIP: Refactor get service to Refactor get service 2023-03-31 14:48:26 +00:00
dstepanov-yadro requested review from storage-core-committers 2023-03-31 14:48:33 +00:00
dstepanov-yadro requested review from storage-core-developers 2023-03-31 14:48:33 +00:00
fyrchik reviewed 2023-04-03 08:39:10 +00:00
pkg/services/object/get/exec.go Outdated
@ -81,4 +78,2 @@
}
func (exec execCtx) context() context.Context {
return exec.ctx
fyrchik commented 2023-04-03 08:38:43 +00:00
Owner

Do we still need the field in the struct?

Do we still need the field in the struct?
dstepanov-yadro commented 2023-04-03 09:28:34 +00:00
Poster
Collaborator

No, we dont. It's dropped.

No, we dont. It's dropped.
fyrchik marked this conversation as resolved
pkg/services/object/get/v2/util.go Outdated
@ -193,2 +59,2 @@
}
}
forwarder := &getRequestForwarder{
OnceResign: &sync.Once{},
fyrchik commented 2023-04-03 08:37:44 +00:00
Owner

Is there any reason you don't just use non-pointer field in struct and omit the initialization?

Is there any reason you don't just use non-pointer field in struct and omit the initialization?
dstepanov-yadro commented 2023-04-03 09:56:28 +00:00
Poster
Collaborator

This is the easiest way to meet this requirement: Values containing the types defined in this package should not be copied. (https://pkg.go.dev/sync)

This is the easiest way to meet this requirement: *Values containing the types defined in this package should not be copied.* (https://pkg.go.dev/sync)
carpawell commented 2023-04-03 15:57:26 +00:00
Collaborator

but forwarder is already a pointer? i am ok with that, just want to come to some approach that will be commonly adopted in that repo (the other code do not use pointer inside a pointer i guess)

but `forwarder` is already a pointer? i am ok with that, just want to come to some approach that will be commonly adopted in that repo (the other code do not use pointer inside a pointer i guess)
dstepanov-yadro commented 2023-04-03 17:22:37 +00:00
Poster
Collaborator

the other code do not use pointer inside a pointer i guess

But you should always think how holder of sync.Mutex (RWMutex, Once...) will be used.

> the other code do not use pointer inside a pointer i guess But you should always think how holder of ```sync.Mutex (RWMutex, Once...)``` will be used.
fyrchik approved these changes 2023-04-03 08:39:24 +00:00
fyrchik approved these changes 2023-04-03 09:49:06 +00:00
aarifullin approved these changes 2023-04-03 13:07:16 +00:00
acid-ant approved these changes 2023-04-03 15:42:59 +00:00
carpawell requested changes 2023-04-03 15:56:00 +00:00
carpawell left a comment
Collaborator

Periods in commits' bodies, please. It is a completed sentence (and quite often more than one sentence).

Periods in commits' bodies, please. It is a completed sentence (and quite often more than one sentence).
pkg/services/object/get/v2/get_forwarder.go Outdated
@ -0,0 +29,4 @@
Stream *streamObjectWriter
}
func (f *getRequestForwarder) forwardRequestToNode(ctx context.Context, addr network.Address, c client.MultiAddressClient, pubkey []byte) (*object.Object, error) {
carpawell commented 2023-04-03 15:50:40 +00:00
Collaborator

so not used second return value?

so not used second return value?
dstepanov-yadro commented 2023-04-03 16:54:37 +00:00
Poster
Collaborator

forwardRequestToNode is used as an argument for func groupAddressRequestForwarder(...) function. groupAddressRequestForwarder requires two values to be returned. It is possible to use a lambda function, but I don't like it.

Do you think that with a lambda and one return value it will be easier to understand?

```forwardRequestToNode``` is used as an argument for ```func groupAddressRequestForwarder(...)``` function. ```groupAddressRequestForwarder``` requires two values to be returned. It is possible to use a lambda function, but I don't like it. Do you think that with a lambda and one return value it will be easier to understand?
carpawell commented 2023-04-05 07:33:56 +00:00
Collaborator

Do you think that with a lambda and one return value it will be easier to understand?

no, do not like lambdas usually as more complex thing for a reader.

is used as an argument for func groupAddressRequestForwarder(...) function

originally, comment was about that: seems like anything that is ...forwarder... should not return objects. if groupAddressRequestForwarder is used somewhere else AND that is not a complete refactor and we plan to reorganize it later, im ok with that

> Do you think that with a lambda and one return value it will be easier to understand? no, do not like lambdas usually as more complex thing for a reader. > is used as an argument for func groupAddressRequestForwarder(...) function originally, comment was about that: seems like anything that is `...forwarder...` should not return objects. if `groupAddressRequestForwarder` is used somewhere else AND that is not a complete refactor and we plan to reorganize it later, im ok with that
dstepanov-yadro commented 2023-04-05 08:11:11 +00:00
Poster
Collaborator

HEAD request requires return result from forwarder

HEAD request requires return result from forwarder
pkg/services/object/get/v2/get_forwarder.go Outdated
@ -0,0 +56,4 @@
return nil, err
}
if err := f.readStream(ctx, c, getStream, pubkey); err != nil {
carpawell commented 2023-04-03 15:52:57 +00:00
Collaborator

can we just return f.readStream err if no details are added? (personally, i would add some details via fmt.Errorf)

related to more than just that line

can we just return `f.readStream` err if no details are added? (personally, i would add some details via `fmt.Errorf`) related to more than just that line
dstepanov-yadro commented 2023-04-03 16:57:41 +00:00
Poster
Collaborator

Fixed.

Fixed.
carpawell marked this conversation as resolved
dstepanov-yadro force-pushed refactoring/OBJECT-3610_getsvc from 3a80d96ff5 to 97f87137af 2023-04-03 16:38:22 +00:00 Compare
dstepanov-yadro commented 2023-04-03 16:38:51 +00:00
Poster
Collaborator

Periods in commits' bodies, please. It is a completed sentence (and quite often more than one sentence).

fixed

> Periods in commits' bodies, please. It is a completed sentence (and quite often more than one sentence). fixed
dstepanov-yadro force-pushed refactoring/OBJECT-3610_getsvc from 97f87137af to b091e78063 2023-04-03 16:57:22 +00:00 Compare
dstepanov-yadro reviewed 2023-04-04 10:09:31 +00:00
dstepanov-yadro reviewed 2023-04-04 10:25:43 +00:00
pkg/services/object/get/v2/util.go Outdated
@ -468,4 +252,0 @@
p.SetRequestForwarder(groupAddressRequestForwarder(func(addr network.Address, c client.MultiAddressClient, pubkey []byte) (*object.Object, error) {
var err error
key, err := s.keyStorage.GetKey(nil)
dstepanov-yadro commented 2023-04-04 10:25:43 +00:00
Poster
Collaborator

@fyrchik @carpawell Is it an error that key is requested at every head and get request? For get range request the key is requested once.
Fixed it in the last commit.

@fyrchik @carpawell Is it an error that ```key``` is requested at every head and get request? For get range request the key is requested once. Fixed it in the last commit.
fyrchik commented 2023-04-04 13:24:06 +00:00
Owner

Not an error, it should return the same key each time used.

Not an error, it should return the same key each time used.
👍 1
dstepanov-yadro commented 2023-04-04 13:26:42 +00:00
Poster
Collaborator

Thx, then it can be requested once.

Thx, then it can be requested once.
fyrchik marked this conversation as resolved
fyrchik reviewed 2023-04-05 04:54:52 +00:00
pkg/services/object/get/assemble.go Outdated
@ -11,3 +11,3 @@
)
func (exec *execCtx) assemble() {
func (exec *execCtx) assemble(ctx context.Context) {
fyrchik commented 2023-04-05 04:54:52 +00:00
Owner

Can we use more meaningful commit message for this? Like a sentence you use in the body.

Can we use more meaningful commit message for this? Like a sentence you use in the body.
fyrchik commented 2023-04-05 04:56:17 +00:00
Owner

And same for the last commit, we have 2 commits with identical message doing pretty different things.

And same for the last commit, we have 2 commits with identical message doing pretty different things.
dstepanov-yadro commented 2023-04-05 08:07:29 +00:00
Poster
Collaborator

Fixed

Fixed
dstepanov-yadro force-pushed refactoring/OBJECT-3610_getsvc from fbd35ce39c to 7dbeb09b94 2023-04-05 08:07:20 +00:00 Compare
dstepanov-yadro requested review from carpawell 2023-04-05 08:11:37 +00:00
fyrchik approved these changes 2023-04-05 08:11:39 +00:00
aarifullin approved these changes 2023-04-05 14:07:13 +00:00
fyrchik merged commit c58ab0c369 into master 2023-04-05 14:38:49 +00:00
Sign in to join this conversation.
Reviewers
No reviewers
acid-ant
carpawell
fyrchik
aarifullin
Labels
Clear labels   
P0

Highest   
P1

High   
P2

Medium   
P3

Low   
badger

Badger DB related issues   
frostfs-adm
   
frostfs-cli
   
frostfs-ir
   
frostfs-lens
   
frostfs-node
   
good first issue

Good for newcomers   
triage
  
Infrastructure

Infrastructure, Automation, CI/CD or DevOps related stuff   
blocked

The issue or PR is blocked by something   
bug

Something is not working   
config

Configuration format update or breaking change   
discussion

Talking about something in order to reach a decision or to exchange ideas   
documentation

Documentation related   
duplicate

This issue or pull request already exists   
enhancement

New feature   
go

Language features adoption   
help wanted

Extra attention is needed   
internal
   
invalid

Something is wrong   
kludge

This is a kludge and we know it   
observability

Related to metrics, tracing and logs   
perfomance

Perfomance related   
question

More information is needed   
refactoring

Refactoring   
wontfix

This won't be fixed
No Label
P0
P1
P2
P3
badger
frostfs-adm
frostfs-cli
frostfs-ir
frostfs-lens
frostfs-node
good first issue
triage
Infrastructure
blocked
bug
config
discussion
documentation
duplicate
enhancement
go
help wanted
internal
invalid
kludge
observability
perfomance
question
refactoring
wontfix
Milestone
Clear milestone
No items
No Milestone
Assignees
Clear assignees
No Assignees
5 Participants
Notifications
Due Date
The due date is invalid or out of range. Please use the format 'yyyy-mm-dd'.

No due date set.

Dependencies

No dependencies set.

Reference: TrueCloudLab/frostfs-node#193
There is no content yet.
Delete Branch "dstepanov-yadro/frostfs-node:refactoring/OBJECT-3610_getsvc"

Deleting a branch is permanent. Although the deleted branch may exist for a short time before cleaning up, in most cases it CANNOT be undone. Continue?