in_calyptia_fleet: Unify fleet config file caching across platforms #10826
Conversation
WalkthroughReplaces symlink-based fleet config handling with a ref-file model, adds ref/deref APIs and recursive deletion, refactors base-dir allocation, adds legacy cleanup, atomic ref writes, safer threaded reloads, standardized 64-bit timestamp formatting, and extended logging. Public APIs/signatures remain unchanged. Changes
Sequence Diagram(s)sequenceDiagram
participant Init as in_calyptia_fleet_init
participant FS as Filesystem
participant Ref as RefManager
participant Thread as ReloadThread
Note over Init,FS: Initialization performs legacy cleanup and creates base dir
Init->>FS: cleanup_legacy_config_state()
FS-->>Init: legacy artifacts removed
Init->>FS: generate_base_fleet_directory() → allocated_path
Note over Ref,FS: Resolve refs instead of reading symlinks
Ref->>FS: fleet_config_deref("new"/"cur"/"old")
FS-->>Ref: resolved config path
Note over Ref,FS: Atomic ref update (write tmp -> rename)
Ref->>FS: write temp ref file
Ref->>FS: rename_file(tmp -> ref)
Note over Ref,Thread: Async reload with copied path, safer cleanup on failure
Ref->>Thread: execute_reload(copy_of_resolved_path)
Thread->>FS: read config from copied path
alt success
Thread-->>Ref: reload success
else failure
Thread->>Ref: resume collectors / cleanup
end
Estimated code review effort🎯 4 (Complex) | ⏱️ ~45 minutes Possibly related PRs
Suggested reviewers
Pre-merge checks (2 passed, 1 warning)❌ Failed checks (1 warning)
✅ Passed checks (2 passed)
✨ Finishing Touches
🧪 Generate unit tests
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
|
@coderabbitai review |
✅ Actions performedReview triggered.
|
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 2
Caution
Some comments are outside the diff and can’t be posted inline due to platform limitations.
⚠️ Outside diff range comments (1)
plugins/in_calyptia_fleet/in_calyptia_fleet.c (1)
617-675: Ownership bug: execute_reload frees cfgpath on error; callers also free → double-freeexecute_reload should not free cfgpath; callers already do. Remove all flb_sds_destroy(cfgpath) in execute_reload.
@@ static int execute_reload(struct flb_in_calyptia_fleet_config *ctx, flb_sds_t cfgpath) - if (parse_config_name_timestamp(ctx, cfgpath, &ctx->config_timestamp) != FLB_TRUE) { - flb_sds_destroy(cfgpath); - return FLB_FALSE; - } + if (parse_config_name_timestamp(ctx, cfgpath, &ctx->config_timestamp) != FLB_TRUE) { + return FLB_FALSE; + } @@ - if (reload == NULL) { - flb_sds_destroy(cfgpath); - return FLB_FALSE; - } + if (reload == NULL) { + return FLB_FALSE; + } @@ - if (reload->cfg_path == NULL) { - flb_free(reload); - flb_sds_destroy(cfgpath); - return FLB_FALSE; - } + if (reload->cfg_path == NULL) { + flb_free(reload); + return FLB_FALSE; + } @@ - flb_sds_destroy(cfgpath); - return FLB_FALSE; + return FLB_FALSE;Note: Also remove the earlier free at Line 625 (same reason).
Also applies to: 1990-2000
🧹 Nitpick comments (5)
plugins/in_calyptia_fleet/in_calyptia_fleet.c (5)
58-59: Guard S_ISDIR macro on WindowsAvoid redefinition warnings in environments that already provide S_ISDIR.
-#define S_ISDIR(m) (((m) & S_IFMT) == S_IFDIR) +#ifndef S_ISDIR +#define S_ISDIR(m) (((m) & S_IFMT) == S_IFDIR) +#endif
210-233: Make new fleet_config_ helpers internal or declare them*These helpers have external linkage but no header prototypes. Either make them static (preferred) or declare them in the header to avoid accidental external use.
-flb_sds_t fleet_config_ref_filename(struct flb_in_calyptia_fleet_config *ctx, char *ref_name) +static flb_sds_t fleet_config_ref_filename(struct flb_in_calyptia_fleet_config *ctx, char *ref_name) @@ -flb_sds_t fleet_config_deref(struct flb_in_calyptia_fleet_config *ctx, char *ref_name) +static flb_sds_t fleet_config_deref(struct flb_in_calyptia_fleet_config *ctx, char *ref_name) @@ -int fleet_config_set_ref(struct flb_in_calyptia_fleet_config *ctx, +static int fleet_config_set_ref(struct flb_in_calyptia_fleet_config *ctx,Also applies to: 239-270, 277-316
582-583: Minor: remove unused variable and favor safer copylen is unused. Using snprintf prevents partial copy footguns.
- ssize_t len; @@ - strncpy(realname, cfgpath, sizeof(realname)-1); + (void)snprintf(realname, sizeof(realname), "%s", cfgpath);
1343-1390: delete_dir: type consistency and robustness
- Use size_t for idx to match entry_count.
- Early-return true if directory is already gone after child deletions.
- int idx; + size_t idx; @@ - if (rmdir(path) != 0) { + if (rmdir(path) != 0) { ret = FLB_FALSE; }Optional: treat ENOENT from rmdir/unlink as success to make cleanup idempotent.
1514-1517: Commit flow: minor message and idempotency polish
- Typo: “unable to delete old configuration by” → “by ref”.
- Consider not erroring when old doesn’t exist; it’s harmless.
- if (calyptia_config_delete_by_ref(ctx, "old") == FLB_FALSE) { - flb_plg_error(ctx->ins, "unable to delete old configuration by"); + if (calyptia_config_delete_by_ref(ctx, "old") == FLB_FALSE) { + flb_plg_error(ctx->ins, "unable to delete old configuration by ref"); return FLB_FALSE; }Also applies to: 1520-1524
📜 Review details
Configuration used: CodeRabbit UI
Review profile: CHILL
Plan: Pro
💡 Knowledge Base configuration:
- MCP integration is disabled by default for public repositories
- Jira integration is disabled by default for public repositories
- Linear integration is disabled by default for public repositories
You can enable these sources in your CodeRabbit configuration.
📒 Files selected for processing (2)
plugins/in_calyptia_fleet/in_calyptia_fleet.c(16 hunks)plugins/in_calyptia_fleet/in_calyptia_fleet.h(1 hunks)
⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (26)
- GitHub Check: pr-windows-build / call-build-windows-package (Windows 64bit, x64, x64-windows-static, 3.31.6)
- GitHub Check: pr-windows-build / call-build-windows-package (Windows 32bit, x86, x86-windows-static, 3.31.6)
- GitHub Check: pr-windows-build / call-build-windows-package (Windows 64bit (Arm64), amd64_arm64, -DCMAKE_SYSTEM_NAME=Windows -DCMA...
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SIMD=Off, 3.31.6, gcc, g++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_ARROW=On, 3.31.6, gcc, g++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SIMD=Off, 3.31.6, clang, clang++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SIMD=On, 3.31.6, gcc, g++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_COVERAGE=On, 3.31.6, gcc, g++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SIMD=On, 3.31.6, clang, clang++)
- GitHub Check: run-ubuntu-unit-tests (-DSANITIZE_ADDRESS=On, 3.31.6, clang, clang++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SANITIZE_THREAD=On, 3.31.6, gcc, g++)
- GitHub Check: run-ubuntu-unit-tests (-DSANITIZE_UNDEFINED=On, 3.31.6, clang, clang++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SMALL=On, 3.31.6, gcc, g++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SANITIZE_MEMORY=On, 3.31.6, clang, clang++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_JEMALLOC=Off, 3.31.6, clang, clang++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SANITIZE_MEMORY=On, 3.31.6, gcc, g++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SANITIZE_THREAD=On, 3.31.6, clang, clang++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_JEMALLOC=On, 3.31.6, clang, clang++)
- GitHub Check: run-ubuntu-unit-tests (-DSANITIZE_UNDEFINED=On, 3.31.6, gcc, g++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SMALL=On, 3.31.6, clang, clang++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_JEMALLOC=On, 3.31.6, gcc, g++)
- GitHub Check: run-ubuntu-unit-tests (-DSANITIZE_ADDRESS=On, 3.31.6, gcc, g++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_JEMALLOC=Off, 3.31.6, gcc, g++)
- GitHub Check: PR - fuzzing test
- GitHub Check: pr-compile-centos-7
- GitHub Check: pr-compile-system-libs (-DFLB_PREFER_SYSTEM_LIBS=On, 3.31.6, clang, clang++, ubuntu-24.04, clang-14)
🔇 Additional comments (5)
plugins/in_calyptia_fleet/in_calyptia_fleet.h (1)
75-77: No internal usages of the old macros found
Verified via ripgrep: there are no occurrences ofnew_fleet_config_filename,cur_fleet_config_filename, orold_fleet_config_filenameoutside theirlegacy_*definitions. Safe to proceed with the renames.plugins/in_calyptia_fleet/in_calyptia_fleet.c (4)
321-321: Time format cast fix LGTMSwitch to %lld with explicit cast avoids UB across platforms.
339-349: Ref-based detection and existence checks LGTMTransition from filename macros to deref() reads clean and keeps semantics.
Also applies to: 366-376, 393-403, 476-490, 492-507, 509-523
1055-1056: Log formatting LGTMConsistent %lld for time_t cast.
2325-2330: Legacy cleanup before dir creation: good, but confirm restart scenariosWiping the whole cache on legacy detection is fine. Please confirm docs call this out and tests cover “mixed-state” nodes (some timestamped dirs + legacy files).
Would you like a small test checklist to validate upgrade paths on Linux/Windows?
alecholmes
force-pushed
the
alec/2025-08-27-unify-fleet-config-file-mgmt
branch
from
92d74a7 to
0687187
Compare
| return NULL; | ||
| } | ||
|
|
||
| static int test_config_is_valid(struct flb_in_calyptia_fleet_config *ctx, |
There was a problem hiding this comment.
Not called in practice and not helpful. Removing altogether.
|
|
||
| #endif | ||
|
|
||
| static int cfl_array_qsort_conf_files(const void *arg_a, const void *arg_b) |
There was a problem hiding this comment.
No need to order files any more.
|
@coderabbitai review |
✅ Actions performedReview triggered.
|
There was a problem hiding this comment.
Actionable comments posted: 2
Caution
Some comments are outside the diff and can’t be posted inline due to platform limitations.
⚠️ Outside diff range comments (1)
plugins/in_calyptia_fleet/in_calyptia_fleet.c (1)
584-595: Avoid basename portability pitfallsbasename() is non-standard on Windows. Use PATH_SEPARATOR scanning (as you already do elsewhere) to extract the filename.
- fname = basename(realname); + { + char *sep = strrchr(realname, PATH_SEPARATOR[0]); + fname = sep ? sep + 1 : realname; + }
♻️ Duplicate comments (1)
plugins/in_calyptia_fleet/in_calyptia_fleet.c (1)
1416-1437: Make delete-by-ref idempotent: treat “no extension/matches” as successReturning FLB_FALSE when the referenced path has no extension or when the glob finds nothing propagates failures into commit/rollback flows, even though there’s nothing to delete. This should be a no-op success.
- if (ext == NULL) { - flb_plg_error(ctx->ins, "referenced config path has no extension: %s", config_path); - flb_sds_destroy(config_path); - return FLB_FALSE; - } + if (ext == NULL) { + flb_plg_debug(ctx->ins, "referenced config path has no extension: %s (nothing to delete)", config_path); + flb_sds_destroy(config_path); + return FLB_TRUE; + } @@ - if (confs == NULL) { - flb_plg_warn(ctx->ins, "config glob did not return any files: %s", config_path); - flb_sds_destroy(config_path); - return FLB_FALSE; - } + if (confs == NULL) { + flb_plg_debug(ctx->ins, "config glob returned no files: %s (nothing to delete)", config_path); + flb_sds_destroy(config_path); + return FLB_TRUE; + } @@ - if (unlink(ref_filename) != 0) { + if (unlink(ref_filename) != 0 && errno != ENOENT) { flb_plg_warn(ctx->ins, "unable to delete config ref file: %s", ref_filename); }Also applies to: 1432-1437, 1456-1463
🧹 Nitpick comments (6)
plugins/in_calyptia_fleet/in_calyptia_fleet.c (6)
297-316: Good fix on double-free/use-after-free; add newline/flush check and make write atomicThe ref write path looks solid now. Consider checking fputc/fflush and using write-then-rename for atomicity.
- (void) fputc('\n', ref_file_ptr); - fclose(ref_file_ptr); + if (fputc('\n', ref_file_ptr) == EOF || fflush(ref_file_ptr) != 0) { + flb_plg_error(ctx->ins, "fleet_config_set_ref: failed to finalize reference file: %s", ref_filename); + fclose(ref_file_ptr); + flb_sds_destroy(ref_filename); + return FLB_FALSE; + } + fclose(ref_file_ptr);Alternative (atomic): write to ".ref.tmp" then rename to ".ref".
484-493: Deduplicate exists_ with a tiny helper*Three near-identical functions. Collapse to exists_ref(ctx, "new"/"cur"/"old") for maintainability.
-static int exists_new_fleet_config(struct flb_in_calyptia_fleet_config *ctx) -{ - int ret = FLB_FALSE; - flb_sds_t cfgnewname; - cfgnewname = fleet_config_deref(ctx, "new"); - if (cfgnewname == NULL) { - return FLB_FALSE; - } - ret = access(cfgnewname, F_OK) == 0 ? FLB_TRUE : FLB_FALSE; - flb_sds_destroy(cfgnewname); - return ret; -} +static int exists_ref(struct flb_in_calyptia_fleet_config *ctx, const char *name) +{ + int ret = FLB_FALSE; + flb_sds_t p = fleet_config_deref(ctx, (char *)name); + if (p == NULL) { return FLB_FALSE; } + ret = access(p, F_OK) == 0 ? FLB_TRUE : FLB_FALSE; + flb_sds_destroy(p); + return ret; +}Then replace bodies with return exists_ref(ctx, "new"/"cur"/"old");
Also applies to: 501-510, 517-526
1345-1388: delete_dir: tighten loop types and error handling
- Mixed signed/unsigned in the loop condition; compare size_t to size_t.
- Optionally handle read-only files on Windows (remove FILE_ATTRIBUTE_READONLY before unlink).
- for (idx = 0; idx < ((ssize_t)files->entry_count); idx++) { + for (idx = 0; idx < files->entry_count; idx++) {
1517-1549: Commit path robustness depends on idempotent delete_by_refWith the above change, committing won’t fail when old artifacts are already gone. As a follow-up, consider treating unlink(new_ref_filename) ENOENT as success too.
- unlink(new_ref_filename); + if (unlink(new_ref_filename) != 0 && errno != ENOENT) { + flb_plg_warn(ctx->ins, "unable to delete 'new' ref: %s", new_ref_filename); + }
1559-1594: Rollback path: mirror idempotency and ENOENT handlingSame reasoning as commit; rely on idempotent delete_by_ref and tolerate missing 'old' ref file.
- unlink(old_ref_filename); + if (unlink(old_ref_filename) != 0 && errno != ENOENT) { + flb_plg_warn(ctx->ins, "unable to delete 'old' ref: %s", old_ref_filename); + }
1988-1998: load_fleet_config: small safety nitexecute_reload already duplicates cfg_path; good. Consider verifying the dereferenced path exists before calling execute_reload to avoid noisy logs.
📜 Review details
Configuration used: CodeRabbit UI
Review profile: CHILL
Plan: Pro
💡 Knowledge Base configuration:
- MCP integration is disabled by default for public repositories
- Jira integration is disabled by default for public repositories
- Linear integration is disabled by default for public repositories
You can enable these sources in your CodeRabbit configuration.
📒 Files selected for processing (2)
plugins/in_calyptia_fleet/in_calyptia_fleet.c(16 hunks)plugins/in_calyptia_fleet/in_calyptia_fleet.h(1 hunks)
🚧 Files skipped from review as they are similar to previous changes (1)
- plugins/in_calyptia_fleet/in_calyptia_fleet.h
🔇 Additional comments (3)
plugins/in_calyptia_fleet/in_calyptia_fleet.c (3)
342-352: Ref-based is_ checks look correct*Using dereferenced paths instead of symlink semantics aligns with the new design.
Also applies to: 369-381, 396-408
2324-2329: Legacy cleanup ordering looks goodRunning cleanup before directory creation matches the upgrade plan and prevents stale state from lingering.
1099-1115: Use abbreviated month (%b) and UTC conversion withtimegmReplace in plugins/in_calyptia_fleet/in_calyptia_fleet.c lines 1101–1102:
- flb_strptime(fbit_last_modified, "%a, %d %B %Y %H:%M:%S GMT", &tm_last_modified); - last_modified = mktime(&tm_last_modified.tm); + if (!flb_strptime(fbit_last_modified, "%a, %d %b %Y %H:%M:%S GMT", &tm_last_modified)) { + flb_plg_error(ctx->ins, "invalid Last-Modified header: %s", fbit_last_modified); + ret = -1; + goto client_error; + } + last_modified = timegm(&tm_last_modified.tm);Ensure
flb_compat.h(or another project header) is included so thattimegmis defined (mapped to_mkgmtimeon Windows).
alecholmes
force-pushed
the
alec/2025-08-27-unify-fleet-config-file-mgmt
branch
from
0687187 to
202e8ec
Compare
|
|
||
| ref_file_ptr = fopen(ref_filename, "r"); | ||
| if (ref_file_ptr == NULL) { | ||
| flb_plg_debug(ctx->ins, "fleet_config_deref: unable to open ref file: %s", ref_filename); |
There was a problem hiding this comment.
here I recommend moving flb_errno() as soon as fopen() fails (just to make sure no other call override errno value), putting this as an flb_plg_error() will give the user visibilty
alecholmes
force-pushed
the
alec/2025-08-27-unify-fleet-config-file-mgmt
branch
from
05cc743 to
a391f55
Compare
There was a problem hiding this comment.
Actionable comments posted: 2
Caution
Some comments are outside the diff and can’t be posted inline due to platform limitations.
⚠️ Outside diff range comments (3)
plugins/in_calyptia_fleet/in_calyptia_fleet.c (3)
2532-2532: Potential null-deref: avoid(void) *configThis dereferences config only to cast to void; if NULL, it can crash. Cast the pointer itself.
- (void) *config; + (void) config;
1203-1216: HTTP Last-Modified parsing: use RFC7231 format and handle parse failureThe format should use abbreviated month ("%b"). Also check flb_strptime’s return.
- flb_strptime(fbit_last_modified, "%a, %d %B %Y %H:%M:%S GMT", &tm_last_modified); + if (flb_strptime(fbit_last_modified, "%a, %d %b %Y %H:%M:%S GMT", &tm_last_modified) == NULL) { + flb_plg_warn(ctx->ins, "failed to parse Last-Modified header"); + ret = -1; + goto client_error; + }
2399-2407: Free config_dir SDS in destroy
Under _WIN32 you call flb_sds_create_size()/flb_sds_printf on ctx->config_dir (plugins/in_calyptia_fleet/in_calyptia_fleet.c:2399-2407) but never free it in in_calyptia_fleet_destroy(). Add a flag to struct flb_in_calyptia_fleet_config (e.g.int config_dir_is_sds;), set it to FLB_TRUE immediately after flb_sds_printf, and in in_calyptia_fleet_destroy() do:if (ctx->config_dir_is_sds) { flb_sds_destroy(ctx->config_dir); }
♻️ Duplicate comments (1)
plugins/in_calyptia_fleet/in_calyptia_fleet.c (1)
660-662: basename() breaks MSVC builds; use a portable basenamebasename() is unavailable on MSVC. Use a separator search that handles both '/' and '\'.
- fname = basename(realname); + /* portable basename: handle both '/' and '\\' */ + char *p1 = strrchr(realname, '/'); + char *p2 = strrchr(realname, '\\'); + fname = (p1 > p2 ? p1 : p2); + fname = fname ? fname + 1 : realname;
🧹 Nitpick comments (4)
plugins/in_calyptia_fleet/in_calyptia_fleet.c (4)
394-405: Harden ref write: check fputc and flush before renameEnsure the newline is written and data hit disk before rename; improves crash-safety.
- fputc('\n', temp_file_ptr); - fclose(temp_file_ptr); + if (fputc('\n', temp_file_ptr) == EOF) { + flb_plg_error(ctx->ins, "fleet_config_set_ref: unable to write newline: %s", temp_filename); + fclose(temp_file_ptr); + unlink(temp_filename); + flb_sds_destroy(ref_filename); + flb_sds_destroy(temp_filename); + return FLB_FALSE; + } + fflush(temp_file_ptr); +#ifndef _WIN32 + int fd = fileno(temp_file_ptr); + if (fd >= 0) { (void) fsync(fd); } +#endif + fclose(temp_file_ptr);
1518-1540: Deletion by ref: treat missing extension as a no-opIf the referenced path has no extension, consider it already cleaned; avoid emitting errors.
- if (ext == NULL) { - flb_plg_error(ctx->ins, "referenced config path has no extension: %s", config_path); - flb_sds_destroy(config_path); - return FLB_FALSE; - } + if (ext == NULL) { + flb_plg_debug(ctx->ins, "referenced config path has no extension: %s", config_path); + flb_sds_destroy(config_path); + return FLB_TRUE; + }
1014-1018: Check flb_sds_printf() result for the Config-Version headerAvoid passing a NULL header value on allocation failure.
- flb_sds_printf(&config_version, "%lld", (long long)ctx->config_timestamp); + if (flb_sds_printf(&config_version, "%lld", (long long)ctx->config_timestamp) == NULL) { + flb_sds_destroy(config_version); + flb_http_client_destroy(client); + flb_upstream_conn_release(u_conn); + return NULL; + }
273-304: Optional: avoid PATH_MAX truncation by reading the entire ref fileUsing a fixed-size buffer can still truncate very long paths. Consider stat/fread into a dynamically sized buffer, then trim CRLF.
📜 Review details
Configuration used: CodeRabbit UI
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (2)
plugins/in_calyptia_fleet/in_calyptia_fleet.c(31 hunks)plugins/in_calyptia_fleet/in_calyptia_fleet.h(1 hunks)
🚧 Files skipped from review as they are similar to previous changes (1)
- plugins/in_calyptia_fleet/in_calyptia_fleet.h
🧰 Additional context used
🧬 Code graph analysis (1)
plugins/in_calyptia_fleet/in_calyptia_fleet.c (5)
src/flb_sds.c (5)
flb_sds_create_size(92-95)flb_sds_destroy(389-399)flb_sds_printf(336-387)flb_sds_create(78-90)flb_sds_cat_safe(204-214)src/flb_lib.c (1)
flb_context_get(993-999)include/fluent-bit/flb_mem.h (2)
flb_calloc(84-96)flb_free(126-128)src/flb_input.c (1)
flb_input_collector_resume(2081-2137)lib/cfl/src/cfl_array.c (1)
cfl_array_destroy(51-69)
⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (29)
- GitHub Check: pr-windows-build / call-build-windows-package (Windows 64bit (Arm64), amd64_arm64, -DCMAKE_SYSTEM_NAME=Windows -DCMA...
- GitHub Check: pr-windows-build / call-build-windows-package (Windows 32bit, x86, x86-windows-static, 3.31.6)
- GitHub Check: pr-windows-build / call-build-windows-package (Windows 64bit, x64, x64-windows-static, 3.31.6)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SIMD=On, 3.31.6, clang, clang++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_ARROW=On, 3.31.6, gcc, g++)
- GitHub Check: run-ubuntu-unit-tests (-DSANITIZE_ADDRESS=On, 3.31.6, clang, clang++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SANITIZE_THREAD=On, 3.31.6, gcc, g++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SIMD=Off, 3.31.6, clang, clang++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SANITIZE_THREAD=On, 3.31.6, clang, clang++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SIMD=On, 3.31.6, gcc, g++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SIMD=Off, 3.31.6, gcc, g++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_COVERAGE=On, 3.31.6, gcc, g++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SANITIZE_MEMORY=On, 3.31.6, clang, clang++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SANITIZE_MEMORY=On, 3.31.6, gcc, g++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_JEMALLOC=On, 3.31.6, clang, clang++)
- GitHub Check: run-ubuntu-unit-tests (-DSANITIZE_UNDEFINED=On, 3.31.6, gcc, g++)
- GitHub Check: run-ubuntu-unit-tests (-DSANITIZE_UNDEFINED=On, 3.31.6, clang, clang++)
- GitHub Check: run-ubuntu-unit-tests (-DSANITIZE_ADDRESS=On, 3.31.6, gcc, g++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SMALL=On, 3.31.6, gcc, g++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SMALL=On, 3.31.6, clang, clang++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_JEMALLOC=Off, 3.31.6, gcc, g++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_JEMALLOC=Off, 3.31.6, clang, clang++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_JEMALLOC=On, 3.31.6, gcc, g++)
- GitHub Check: pr-compile-system-libs (-DFLB_PREFER_SYSTEM_LIBS=On, 3.31.6, gcc, g++, ubuntu-22.04, clang-12)
- GitHub Check: pr-compile-system-libs (-DFLB_PREFER_SYSTEM_LIBS=On, 3.31.6, clang, clang++, ubuntu-24.04, clang-14)
- GitHub Check: pr-compile-system-libs (-DFLB_PREFER_SYSTEM_LIBS=On, 3.31.6, clang, clang++, ubuntu-22.04, clang-12)
- GitHub Check: pr-compile-system-libs (-DFLB_PREFER_SYSTEM_LIBS=On, 3.31.6, gcc, g++, ubuntu-24.04, clang-14)
- GitHub Check: PR - fuzzing test
- GitHub Check: pr-compile-centos-7
alecholmes
force-pushed
the
alec/2025-08-27-unify-fleet-config-file-mgmt
branch
from
a391f55 to
2e540e5
Compare
There was a problem hiding this comment.
Actionable comments posted: 0
Caution
Some comments are outside the diff and can’t be posted inline due to platform limitations.
⚠️ Outside diff range comments (3)
plugins/in_calyptia_fleet/in_calyptia_fleet.c (3)
1416-1423: Infinite recursion on Windows globbing (use buf, not path)Calling read_glob_win(path, ...) recurses with the same pattern forever.
Apply:
- if (strchr(p1, '*')) { - if (read_glob_win(path, list) == NULL) { + if (strchr(p1, '*')) { + if (read_glob_win(buf, list) == NULL) { cfl_array_destroy(list); FindClose(hnd); return NULL; } continue; }
1203-1206: HTTP Last-Modified parsing: wrong month token and timezoneRFC dates use abbreviated month (%b), and the value is GMT. mktime() treats it as local time.
Apply:
- flb_strptime(fbit_last_modified, "%a, %d %B %Y %H:%M:%S GMT", &tm_last_modified); - last_modified = mktime(&tm_last_modified.tm); + flb_strptime(fbit_last_modified, "%a, %d %b %Y %H:%M:%S GMT", &tm_last_modified); +#ifdef _WIN32 + last_modified = _mkgmtime(&tm_last_modified.tm); +#else + last_modified = timegm(&tm_last_modified.tm); +#endif + if (last_modified == (time_t)-1) { + goto client_error; + }
2131-2155: Path traversal risk when writing fleet filesname comes from the control plane; it can escape fleetdir via “../”, absolute paths, or separators.
Apply a hard deny-list before composing the path:
static int create_fleet_file(flb_sds_t fleetdir, const char *name, int nlen, const char *b64_content, int blen) { flb_sds_t fname; flb_sds_t dst; size_t dlen = 2 * blen; FILE *fp; int ret; + int i; + + /* Basic path sanitization: reject absolute or traversal */ + if (name == NULL || nlen <= 0) { + return -1; + } + /* absolute paths */ + if (name[0] == '/' || name[0] == '\\') { + return -1; + } + /* traversal and separators */ + for (i = 0; i < nlen; i++) { + if (name[i] == '/' || name[i] == '\\') { + return -1; + } + } + if (strstr(name, "..") != NULL) { + return -1; + }Optional: normalize and verify the final path stays under fleetdir via realpath-equivalent before fopen.
♻️ Duplicate comments (3)
plugins/in_calyptia_fleet/in_calyptia_fleet.c (3)
1519-1525: Deletion-by-ref: treat no-extension as no-op success (not error)Aligns with “no matches” behavior and avoids noisy logs.
Apply:
- if (ext == NULL) { - flb_plg_error(ctx->ins, "referenced config path has no extension: %s", config_path); - flb_sds_destroy(config_path); - return FLB_FALSE; - } + if (ext == NULL) { + flb_plg_debug(ctx->ins, "referenced config path has no extension: %s", config_path); + flb_sds_destroy(config_path); + return FLB_TRUE; + }
299-304: Trim CRLF, not just LF, when parsing ref filesOn Windows, a stray '\r' can sneak into the path and break subsequent file ops.
Apply:
- /* Remove trailing newline if present */ - if (len > 0 && line[len - 1] == '\n') { - line[len - 1] = '\0'; - } + /* Remove trailing newline(s), handle CRLF */ + while (len > 0 && (line[len - 1] == '\n' || line[len - 1] == '\r')) { + line[--len] = '\0'; + }
660-663: basename() is non-portable; avoid MSVC breakbasename isn’t available on MSVC. Use a portable tail-extraction.
Apply:
- snprintf(realname, sizeof(realname), "%s", cfgpath); - fname = basename(realname); + snprintf(realname, sizeof(realname), "%s", cfgpath); + fname = strrchr(realname, PATH_SEPARATOR[0]); + fname = fname ? fname + 1 : realname;
🧹 Nitpick comments (1)
plugins/in_calyptia_fleet/in_calyptia_fleet.c (1)
375-401: Open ref temp file in binary mode; ensure durability before rename“w”/text can translate newlines on Windows; binary is safer. Optionally fsync dir on POSIX if durability matters.
Apply:
- temp_file_ptr = fopen(temp_filename, "w"); + temp_file_ptr = fopen(temp_filename, "wb"); @@ - fputc('\n', temp_file_ptr); + fputc('\n', temp_file_ptr); fclose(temp_file_ptr);If you need crash-safety across power loss, add fsync on the file and its parent dir before/after rename on POSIX.
📜 Review details
Configuration used: CodeRabbit UI
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (2)
plugins/in_calyptia_fleet/in_calyptia_fleet.c(31 hunks)plugins/in_calyptia_fleet/in_calyptia_fleet.h(1 hunks)
🚧 Files skipped from review as they are similar to previous changes (1)
- plugins/in_calyptia_fleet/in_calyptia_fleet.h
🧰 Additional context used
🧬 Code graph analysis (1)
plugins/in_calyptia_fleet/in_calyptia_fleet.c (5)
src/flb_sds.c (5)
flb_sds_create_size(92-95)flb_sds_destroy(389-399)flb_sds_printf(336-387)flb_sds_create(78-90)flb_sds_cat_safe(204-214)src/flb_lib.c (1)
flb_context_get(993-999)include/fluent-bit/flb_mem.h (2)
flb_calloc(84-96)flb_free(126-128)src/flb_input.c (1)
flb_input_collector_resume(2081-2137)lib/cfl/src/cfl_array.c (1)
cfl_array_destroy(51-69)
⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (29)
- GitHub Check: pr-windows-build / call-build-windows-package (Windows 64bit (Arm64), amd64_arm64, -DCMAKE_SYSTEM_NAME=Windows -DCMA...
- GitHub Check: pr-windows-build / call-build-windows-package (Windows 32bit, x86, x86-windows-static, 3.31.6)
- GitHub Check: pr-windows-build / call-build-windows-package (Windows 64bit, x64, x64-windows-static, 3.31.6)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SIMD=On, 3.31.6, clang, clang++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_ARROW=On, 3.31.6, gcc, g++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SANITIZE_THREAD=On, 3.31.6, clang, clang++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SIMD=Off, 3.31.6, clang, clang++)
- GitHub Check: run-ubuntu-unit-tests (-DSANITIZE_UNDEFINED=On, 3.31.6, clang, clang++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SIMD=On, 3.31.6, gcc, g++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SIMD=Off, 3.31.6, gcc, g++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SANITIZE_MEMORY=On, 3.31.6, clang, clang++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SANITIZE_THREAD=On, 3.31.6, gcc, g++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_COVERAGE=On, 3.31.6, gcc, g++)
- GitHub Check: run-ubuntu-unit-tests (-DSANITIZE_ADDRESS=On, 3.31.6, clang, clang++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SANITIZE_MEMORY=On, 3.31.6, gcc, g++)
- GitHub Check: run-ubuntu-unit-tests (-DSANITIZE_UNDEFINED=On, 3.31.6, gcc, g++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_JEMALLOC=On, 3.31.6, gcc, g++)
- GitHub Check: run-ubuntu-unit-tests (-DSANITIZE_ADDRESS=On, 3.31.6, gcc, g++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SMALL=On, 3.31.6, clang, clang++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SMALL=On, 3.31.6, gcc, g++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_JEMALLOC=Off, 3.31.6, gcc, g++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_JEMALLOC=Off, 3.31.6, clang, clang++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_JEMALLOC=On, 3.31.6, clang, clang++)
- GitHub Check: pr-compile-system-libs (-DFLB_PREFER_SYSTEM_LIBS=On, 3.31.6, gcc, g++, ubuntu-24.04, clang-14)
- GitHub Check: pr-compile-centos-7
- GitHub Check: pr-compile-system-libs (-DFLB_PREFER_SYSTEM_LIBS=On, 3.31.6, gcc, g++, ubuntu-22.04, clang-12)
- GitHub Check: pr-compile-system-libs (-DFLB_PREFER_SYSTEM_LIBS=On, 3.31.6, clang, clang++, ubuntu-24.04, clang-14)
- GitHub Check: pr-compile-system-libs (-DFLB_PREFER_SYSTEM_LIBS=On, 3.31.6, clang, clang++, ubuntu-22.04, clang-12)
- GitHub Check: PR - fuzzing test
🔇 Additional comments (5)
plugins/in_calyptia_fleet/in_calyptia_fleet.c (5)
155-202: Directory builder looks solidNull checks and early frees are correct; SDS lifetime is handled properly.
316-330: Cross-platform replace-atomic rename looks goodWindows uses MoveFileExA with REPLACE_EXISTING|WRITE_THROUGH; POSIX uses rename(). Nice.
1108-1167: Timestamp-newness check: LGTMClear logic, correct cleanup on all paths, and ignores non-timestamped files.
1612-1697: Commit/rollback flows are coherentRef updates and cleanups are ordered correctly; errors are surfaced.
2441-2444: Legacy cleanup should be best-effort (already is)Warn-and-continue is appropriate here.
alecholmes
force-pushed
the
alec/2025-08-27-unify-fleet-config-file-mgmt
branch
from
2e540e5 to
38b11b7
Compare
Signed-off-by: Alec Holmes <alec@chronosphere.io>
alecholmes
force-pushed
the
alec/2025-08-27-unify-fleet-config-file-mgmt
branch
from
38b11b7 to
7883904
Compare
There was a problem hiding this comment.
Actionable comments posted: 0
Caution
Some comments are outside the diff and can’t be posted inline due to platform limitations.
⚠️ Outside diff range comments (1)
plugins/in_calyptia_fleet/in_calyptia_fleet.c (1)
1259-1261: Guard against uninitialized last_modifiedWhen dst != NULL, last_modified isn’t set; writing it via time_last_modified stores garbage. Only assign when computed.
- if (time_last_modified) { - *time_last_modified = last_modified; - } + if (time_last_modified && dst == NULL) { + *time_last_modified = last_modified; + }
♻️ Duplicate comments (2)
plugins/in_calyptia_fleet/in_calyptia_fleet.c (2)
664-666: Windows build break: replace non-portable basename()basename() isn’t available on MSVC. Use a portable separator search (handle both separators so cross-OS paths don’t choke).
- fname = basename(realname); + /* Portable basename: prefer current-OS separator, but handle the other too */ + fname = strrchr(realname, PATH_SEPARATOR[0]); +#ifdef _WIN32 + { + char *alt = strrchr(realname, '/'); + if (alt && (!fname || alt > fname)) { fname = alt; } + } +#else + { + char *alt = strrchr(realname, '\\'); + if (alt && (!fname || alt > fname)) { fname = alt; } + } +#endif + fname = fname ? fname + 1 : realname;
303-307: Fix Windows path parsing: strip CRLF when reading ref filesRef files are written in text mode; on Windows they contain CRLF. Only trimming '\n' leaves a trailing '\r', breaking path lookups (access/stat/unlink). Trim both.
Apply:
- /* Remove trailing newline if present */ - if (len > 0 && line[len - 1] == '\n') { - line[len - 1] = '\0'; - } + /* Remove trailing newline(s), handle CRLF */ + while (len > 0 && (line[len - 1] == '\n' || line[len - 1] == '\r')) { + line[--len] = '\0'; + }
🧹 Nitpick comments (4)
plugins/in_calyptia_fleet/in_calyptia_fleet.c (4)
273-301: Avoid PATH_MAX truncation when dereferencing ref filesUsing a fixed PATH_MAX buffer plus “newline required” check can reject valid long paths and older ref files lacking a trailing newline. Consider reading the whole file (fseek/ftell + fread) and accepting EOF-terminated lines.
398-399: Check newline write result when updating refTiny robustness win: verify fputc succeeds before renaming.
- fputc('\n', temp_file_ptr); + if (fputc('\n', temp_file_ptr) == EOF) { + flb_plg_error(ctx->ins, "fleet_config_set_ref: unable to write newline: %s", temp_filename); + fclose(temp_file_ptr); + unlink(temp_filename); + flb_sds_destroy(ref_filename); + flb_sds_destroy(temp_filename); + return FLB_FALSE; + }
1451-1496: Recursive delete: consider soft-failing on partial cleanupFunction returns FLB_FALSE on any single stat/unlink failure, but still best-effort deletes most entries. If callers treat FLB_FALSE as fatal, this could block upgrades due to a single stubborn file. Consider logging and continuing, and only fail if the root rmdir fails.
1647-1656: commit(): unlinking “new.ref” without checking errors is fine, but log errno on failureOptional: add flb_errno() before warn to aid troubleshooting.
📜 Review details
Configuration used: CodeRabbit UI
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (2)
plugins/in_calyptia_fleet/in_calyptia_fleet.c(32 hunks)plugins/in_calyptia_fleet/in_calyptia_fleet.h(1 hunks)
🚧 Files skipped from review as they are similar to previous changes (1)
- plugins/in_calyptia_fleet/in_calyptia_fleet.h
🧰 Additional context used
🧬 Code graph analysis (1)
plugins/in_calyptia_fleet/in_calyptia_fleet.c (4)
src/flb_sds.c (5)
flb_sds_create_size(92-95)flb_sds_destroy(389-399)flb_sds_printf(336-387)flb_sds_create(78-90)flb_sds_cat_safe(204-214)include/fluent-bit/flb_mem.h (2)
flb_calloc(84-96)flb_free(126-128)src/flb_input.c (1)
flb_input_collector_resume(2081-2137)lib/cfl/src/cfl_array.c (1)
cfl_array_destroy(51-69)
⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (29)
- GitHub Check: pr-windows-build / call-build-windows-package (Windows 32bit, x86, x86-windows-static, 3.31.6)
- GitHub Check: pr-windows-build / call-build-windows-package (Windows 64bit (Arm64), amd64_arm64, -DCMAKE_SYSTEM_NAME=Windows -DCMA...
- GitHub Check: pr-windows-build / call-build-windows-package (Windows 64bit, x64, x64-windows-static, 3.31.6)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SIMD=On, 3.31.6, gcc, g++)
- GitHub Check: pr-compile-system-libs (-DFLB_PREFER_SYSTEM_LIBS=On, 3.31.6, clang, clang++, ubuntu-22.04, clang-12)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SIMD=On, 3.31.6, clang, clang++)
- GitHub Check: pr-compile-system-libs (-DFLB_PREFER_SYSTEM_LIBS=On, 3.31.6, clang, clang++, ubuntu-24.04, clang-14)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SIMD=Off, 3.31.6, clang, clang++)
- GitHub Check: pr-compile-system-libs (-DFLB_PREFER_SYSTEM_LIBS=On, 3.31.6, gcc, g++, ubuntu-24.04, clang-14)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_ARROW=On, 3.31.6, gcc, g++)
- GitHub Check: pr-compile-system-libs (-DFLB_PREFER_SYSTEM_LIBS=On, 3.31.6, gcc, g++, ubuntu-22.04, clang-12)
- GitHub Check: run-ubuntu-unit-tests (-DSANITIZE_ADDRESS=On, 3.31.6, clang, clang++)
- GitHub Check: run-ubuntu-unit-tests (-DSANITIZE_ADDRESS=On, 3.31.6, gcc, g++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SIMD=Off, 3.31.6, gcc, g++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SMALL=On, 3.31.6, clang, clang++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SANITIZE_THREAD=On, 3.31.6, clang, clang++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_COVERAGE=On, 3.31.6, gcc, g++)
- GitHub Check: PR - fuzzing test
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SANITIZE_MEMORY=On, 3.31.6, gcc, g++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SANITIZE_MEMORY=On, 3.31.6, clang, clang++)
- GitHub Check: run-ubuntu-unit-tests (-DSANITIZE_UNDEFINED=On, 3.31.6, clang, clang++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SANITIZE_THREAD=On, 3.31.6, gcc, g++)
- GitHub Check: run-ubuntu-unit-tests (-DSANITIZE_UNDEFINED=On, 3.31.6, gcc, g++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_JEMALLOC=On, 3.31.6, clang, clang++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_JEMALLOC=Off, 3.31.6, gcc, g++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_JEMALLOC=Off, 3.31.6, clang, clang++)
- GitHub Check: run-ubuntu-unit-tests (-DFLB_JEMALLOC=On, 3.31.6, gcc, g++)
- GitHub Check: pr-compile-centos-7
- GitHub Check: run-ubuntu-unit-tests (-DFLB_SMALL=On, 3.31.6, gcc, g++)
🔇 Additional comments (4)
plugins/in_calyptia_fleet/in_calyptia_fleet.c (4)
320-334: Cross-platform atomic replace looks goodSwitch to MoveFileExA with REPLACE_EXISTING/WRITE_THROUGH on Windows and rename() elsewhere is correct.
750-767: Threaded reload error handling improvements: LGTMGood: checks for pthread_attr_init/setdetachstate/create, consistent cleanup, and collector resume on failure.
Also applies to: 770-777
1120-1171: Timestamp freshness check logic: LGTM (nice addition)Clean globbing and comparison; returns true when directory is empty. Reads well.
2444-2449: Legacy cleanup on init: LGTMGracefully soft-failing and proceeding is the right call to avoid blocking startup.
4 participants
Overview
The fleet feature locally caches user config files fetched from the control plane. To support hot reloading and rollback when configs change, the implementation may cache multiple versions of the config at any given time. Prior to this PR, there are effectively separate implementations for POSIX and Windows, which causes a few problems:
This PR refactors how the implementation tracks the various versions of configs, unifying around a single approach that works for both POSIX and Windows. This replaces symlinks (on POSIX) and assumptions around file ordering (on Windows) with a single approach of approximating symlink behavior in a cross-platform way. This change was designed to safely roll forward and rollback from older fluent-bit binaries.
I have tested this on both Linux and Windows.
Eng design
It is useful to explicitly track which config files are the stable and which ones are new. Fleet does this today on POSIX systems using symlinks (
new.confis just fetched and maybe invalid,old.confis the last stable config):Windows does not have symlinks so it instead implicitly relies on the number of files and their lexical order, e.g. it assumes the last 3 files in the config directory are the new, unstable configs.
This new approach approximates symlinks without using them. Instead of symlinks, it uses files that contain the path of the relevant config:
In the example above, the contents of the
new.refwould be/tmp/calyptia-fleet/4803d408a0d43a9281ef4ae7720dfd7a23003204d98a4a4ffedca2f036db7206/hello/1755308 054.conf.Binary rollout and rollback
We don't want failures or incorrect behavior when a customer upgrades their fluent-bit binary to include this change, or downgrades to an older binary. Incorrect behavior either is failure to work at all, or incorrectly using an older config.
Upgrade path:
Downgrade path:
*.reffiles are ignored (since the code doesn't know about them). The latest config will be fetched and cached, and symlinks recreated.Enter
[N/A]in the box, if an item is not applicable to your change.Testing
Before we can approve your change; please submit the following in a comment:
If this is a change to packaging of containers or native binaries then please confirm it works for all targets.
ok-package-testlabel to test for all targets (requires maintainer to do).Documentation
Backporting
Fluent Bit is licensed under Apache 2.0, by submitting this pull request I understand that this code will be released under the terms of that license.
Summary by CodeRabbit
New Features
Bug Fixes
Refactor
Chores