TrueCloudLab/frostfs-node
21
1
Fork 27
Code Issues 108 Pull requests 3 Projects Releases 1 Activity Actions

container/ape: Ignore an error when getting a role #1454

Merged
fyrchik merged 1 commit from a-savchuk/frostfs-node:container-get-on-first-epoch into master 2024-10-30 10:41:29 +00:00
Conversation 0 Commits 1 Files changed 1 +9 -12
a-savchuk commented 2024-10-29 13:11:58 +00:00
Member
Copy link

Close #1448

When getting a role in the APE checker for the container services, an error may be returned if network maps of the previous two epochs don't have enough nodes to fulfil a container placement policy. It's a logical error, so we should ignore it.

The affected function isContainerNode is used for determining if an actor is a container role when checking an APE rule. It's guaranteed that the policy of a container is valid because that container has been with that policy before. So, the only error could be returned if the map doesn't have enough nodes. It's a logical error, we should ignore it because it doesn't affect an actor role determining.

Before:

~/repos/frostfs/dev-env$ ./frostfs-cli -r s01.frostfs.devenv:8080 container get \
    -g --cid G1B2X97U5w6iLbjazJUNgkaA4WQ9LDvDnw1vyS3QyfKv
rpc error: status: code = 1024 message = not enough nodes to SELECT from

After:

~/repos/frostfs/dev-env$ ./frostfs-cli -r s01.frostfs.devenv:8080 container get \
    -g --cid 9HnHVeeNDMz8Zbphio4APTKQnVf7Xbremw1x8BaCE6NB
CID: 9HnHVeeNDMz8Zbphio4APTKQnVf7Xbremw1x8BaCE6NB
owner ID: NbUgTSFvPmsRxmGeWpuuGeJUoRoi6PErcM
basic ACL: 1c8c8ccc (private)
       RangeHASH    Range      Search     Delete     Put        Head       Get
0 1    1 1 0 0      1 0 0 0    1 1 0 0    1 0 0 0    1 1 0 0    1 1 0 0    1 1 0 0
X F    U S O B      U S O B    U S O B    U S O B    U S O B    U S O B    U S O B
  X-Sticky F-Final U-User S-System O-Others B-Bearer
created: 2024-10-29 17:35:23 +0300 MSK
attributes:
        Timestamp=1730212523
placement policy:
REP 1
CBF 1
Close #1448 When getting a role in the APE checker for the container services, an error may be returned if network maps of the previous two epochs don't have enough nodes to fulfil a container placement policy. It's a logical error, so we should ignore it. The affected function `isContainerNode` is used for determining if an actor is a container role when checking an APE rule. It's guaranteed that the policy of a container is valid because that container has been with that policy before. So, the only error could be returned if the map doesn't have enough nodes. It's a logical error, we should ignore it because it doesn't affect an actor role determining. Before: ``` ~/repos/frostfs/dev-env$ ./frostfs-cli -r s01.frostfs.devenv:8080 container get \ -g --cid G1B2X97U5w6iLbjazJUNgkaA4WQ9LDvDnw1vyS3QyfKv rpc error: status: code = 1024 message = not enough nodes to SELECT from ``` After: ``` ~/repos/frostfs/dev-env$ ./frostfs-cli -r s01.frostfs.devenv:8080 container get \ -g --cid 9HnHVeeNDMz8Zbphio4APTKQnVf7Xbremw1x8BaCE6NB CID: 9HnHVeeNDMz8Zbphio4APTKQnVf7Xbremw1x8BaCE6NB owner ID: NbUgTSFvPmsRxmGeWpuuGeJUoRoi6PErcM basic ACL: 1c8c8ccc (private) RangeHASH Range Search Delete Put Head Get 0 1 1 1 0 0 1 0 0 0 1 1 0 0 1 0 0 0 1 1 0 0 1 1 0 0 1 1 0 0 X F U S O B U S O B U S O B U S O B U S O B U S O B U S O B X-Sticky F-Final U-User S-System O-Others B-Bearer created: 2024-10-29 17:35:23 +0300 MSK attributes: Timestamp=1730212523 placement policy: REP 1 CBF 1 ```
a-savchuk changed title from WIP: services/container/ape: Ignore an error when getting a role to WIP: container/ape: Ignore an error when getting a role 2024-10-29 13:13:44 +00:00
a-savchuk force-pushed container-get-on-first-epoch from 96b5b6a28b to 2fa82acd1c 2024-10-29 13:13:48 +00:00 Compare
a-savchuk force-pushed container-get-on-first-epoch from 2fa82acd1c to e70c2cedbe 2024-10-29 14:36:02 +00:00 Compare
a-savchuk changed title from WIP: container/ape: Ignore an error when getting a role to container/ape: Ignore an error when getting a role 2024-10-29 14:41:45 +00:00
a-savchuk requested review from storage-core-developers 2024-10-29 14:41:56 +00:00
a-savchuk requested review from storage-core-committers 2024-10-29 14:41:59 +00:00
fyrchik approved these changes 2024-10-29 14:56:39 +00:00
Dismissed
dstepanov-yadro reviewed 2024-10-29 15:18:32 +00:00
pkg/services/container/ape.go Outdated
@ -562,0 +554,4 @@
// don't have enough nodes to fulfil a container placement policy.
// It's a logical error, so we should ignore it.
// See https://git.frostfs.info/TrueCloudLab/frostfs-node/issues/1448.
cnrVectors, _ := nm.ContainerNodes(cont.Value.PlacementPolicy(), binCnrID)
dstepanov-yadro commented 2024-10-29 15:18:32 +00:00
Member
Copy link

Why is any error ignored, but not only specific one?

Why is any error ignored, but not only specific one?
a-savchuk commented 2024-10-30 09:43:34 +00:00
Author
Member
Copy link

Made the PR description more detailed:

The affected function isContainerNode is used for determining if an actor is a container role when checking an APE rule. It's guaranteed that the policy of a container is valid because that container has been with that policy before. So, the only error could be returned if the map doesn't have enough nodes. It's a logical error, we should ignore it because it doesn't affect an actor role determining.

Moreover, that error is not a part of the SDK interface, so if I ignore only that error, it'll look like this

if err != nil && !strings.Contains(err.Error(), "not enough nodes to SELECT from") {
    return false, err
}
Made the PR description more detailed: > The affected function `isContainerNode` is used for determining if an actor is a container role when checking an APE rule. It's guaranteed that the policy of a container is valid because that container has been with that policy before. So, the only error could be returned if the map doesn't have enough nodes. It's a logical error, we should ignore it because it doesn't affect an actor role determining. Moreover, that error is not a part of the SDK interface, so if I ignore only that error, it'll look like this ```go if err != nil && !strings.Contains(err.Error(), "not enough nodes to SELECT from") { return false, err } ```
dstepanov-yadro marked this conversation as resolved
acid-ant reviewed 2024-10-29 15:31:14 +00:00
pkg/services/container/ape.go Outdated
@ -562,0 +553,4 @@
// An error may be returned if network maps of the previous two epochs
// don't have enough nodes to fulfil a container placement policy.
// It's a logical error, so we should ignore it.
// See https://git.frostfs.info/TrueCloudLab/frostfs-node/issues/1448.
acid-ant commented 2024-10-29 15:31:14 +00:00
Member
Copy link

Link to the issue looks redundant here, it will be in the commit message.

Link to the issue looks redundant here, it will be in the commit message.
a-savchuk commented 2024-10-30 09:49:40 +00:00
Author
Member
Copy link

Yeah, removed it✅

Yeah, removed it✅
a-savchuk changed title from container/ape: Ignore an error when getting a role to WIP: container/ape: Ignore an error when getting a role 2024-10-30 09:43:58 +00:00
a-savchuk dismissed fyrchik's review 2024-10-30 09:48:47 +00:00
Reason:

New commits pushed, approval review dismissed automatically according to repository settings

a-savchuk force-pushed container-get-on-first-epoch from 6daa1fda7f to d28a5d2d7a 2024-10-30 09:52:13 +00:00 Compare
dstepanov-yadro approved these changes 2024-10-30 09:53:12 +00:00
a-savchuk changed title from WIP: container/ape: Ignore an error when getting a role to container/ape: Ignore an error when getting a role 2024-10-30 09:54:12 +00:00
fyrchik merged commit d28a5d2d7a into master 2024-10-30 10:41:29 +00:00
a-savchuk deleted branch container-get-on-first-epoch 2024-10-30 11:10:57 +00:00
Sign in to join this conversation.
Reviewers
No reviewers
dstepanov-yadro
TrueCloudLab/storage-core-developers
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 labels
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
Projects
Clear projects
No items
No project
Assignees
Clear assignees
No assignees
4 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#1454
No description provided.
Delete branch "a-savchuk/frostfs-node:container-get-on-first-epoch"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?