Comments (10)
@ntsh-gpt I deleted that comment, was premature, #184 won't solve it.
from raft.
Thanks, we'll investigate.
from raft.
This looks like a bug that indeed should be investigated. However, it should be greatly mitigated if you implement a graceful shutdown in your code, by catching whatever signal you are sending to the process, then calling raft_close
and waiting for its callback to fire before exiting the process. That's exactly what is done in the example code.
from raft.
@freeekanayaka I am following the same approach as mentioned in the example code.
However I would like to point out exact case for the issue,
Working :
struct A {
int64_t x;
int64_t y;
};
struct A data;
raft_buffer buf;
buf.len = sizeof(struct A);
buf.base = raft_malloc(buf.len);
// assign data values to buf.base
raft_apply(&raft, req, &buf, 1, serverApplyCb)
Not working :
FILE* f = fopen("path");
buf.len = ftell(f) ;
fseek(f, 0, SEEK_SET);
buf.base = raft_malloc(buf.len);
fread(buf.base,1,buf.len,f);
fclose(f);
raft_apply(&raft, req, &buf, 1, serverApplyCb)
Note : I tried memcpy a char array to buf.base, still same issue.
from raft.
I believe I'm also hitting this bug. The idea of calling raft_close is useful, thanks, but this is a bit worrisome, and calls into question whether I should rely on this implementation. Raft is supposed to help with durability of log entries. That means that if my server dies due to power failure, I should be able to gracefully restore my machine. What is raft_close doing that can't be done after a raft_apply to make sure that the open segment can be read again? I'm happy to contribute to make things work, but I want to make sure that such a use case (power failure) is supposed to be supported.
from raft.
I believe I'm also hitting this bug. The idea of calling raft_close is useful, thanks, but this is a bit worrisome, and calls into question whether I should rely on this implementation. Raft is supposed to help with durability of log entries. That means that if my server dies due to power failure, I should be able to gracefully restore my machine. What is raft_close doing that can't be done after a raft_apply to make sure that the open segment can be read again? I'm happy to contribute to make things work, but I want to make sure that such a use case (power failure) is supposed to be supported.
We do have Jepsen tests that simulate hard process crashes (SIGKILL) and recovery after that, probably we should run those tests more times and see if we can reproduce this particular bug (or just try what @ntsh-gpt as pointed).
As I said, this really looks like a bug that should be fixed, so the suggestion of using raft_close
is just a mitigation, I'm not meaning that it is supposed to be a required procedure in order for the implementation to work.
from raft.
@MathieuBordere Will try out with #184.
from raft.
@ntsh-gpt I deleted that comment, was premature, #184 won't solve it.
Yeah, #184 should only be relevant for snapshots. Closed segments should alread be atomically created (by renaming an open segment).
from raft.
@ntsh-gpt Are you hitting the same issue as in #189, if yes, can I close?
from raft.
Closing for inactivity, assuming that the fix above took care of it.
If not, let us know and we'll re-open.
from raft.
Related Issues (20)
- missing getrandom on centos 7 HOT 1
- [question] Usage of raft_add, raft_remove, and raft_assign HOT 2
- [question] When is the first leader elected? HOT 2
- recvAppendEntries: Assertion `r->state == RAFT_FOLLOWER || r->state == RAFT_CANDIDATE` failed HOT 9
- [question] Forwarding request to leader HOT 8
- Potential use-after-free in handling of raft_transfer HOT 5
- Fix the 32-bit CI HOT 1
- can't build on m1 mac (`<linux/xxxxx.h>` is missing) HOT 3
- v1.x RFC: pull based approach HOT 5
- raft_start(): io: closed segment xxx is past last snapshot xxx HOT 4
- [question] Understanding how to add new servers HOT 15
- ./configure fails when no external dependency is found HOT 4
- Assertion: src/uv_truncate.c:168: UvTruncate: Assertion `index < uv->append_next_index' failed. HOT 3
- src/replication.c:457: getRequest: Assertion `req->type == type' failed. HOT 3
- Segment writes blocked when taking a snapshot. HOT 3
- src/log.c:87: refsTryInsert: Assertion `next_slot->term != term' failed. HOT 9
- Jepsen: Another truncate-related assertion failure
- [suggestion] add lock/unlock function pointers to raft_io (and use them) HOT 1
- `uvOsFallocateEmulation` taking too long time HOT 1
- CI: `xfs` test failures HOT 1
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from raft.