Add native CDC support by tclinkenbeard-oai · Pull Request #13287 · apple/foundationdb

tclinkenbeard-oai · 2026-05-27T16:58:20Z

This PR extends FDB to support natively streaming mutations to tracked key ranges. Clients register CDC streams and these are tracked via CDC proxies, which peek from the log system and pop once acknowledgements are received from clients. New pseudo-tags are added to mutations by commit proxies in order to support this. Overlapping CDC streams are supported, and a mutation can't be safely popped until all CDC streams have sent corresponding acknowledgements. This PR does not add support for the CDC interface in the bindings, that is planned for a future PR.

Full documentation of the feature is added in design/cdc.md.

100k/100k filtered *NativeCDC* correctness tests passed.

foundationdb-ci · 2026-05-27T17:02:07Z

Result of foundationdb-pr-clang-arm on Linux CentOS 7

Commit ID: 9b4b2f5
Duration 0:03:36
Result: ❌ FAILED
Error: Error while executing command: if [[ $(git diff --shortstat 2> /dev/null | tail -n1) == "" ]]; then echo "CODE FORMAT CLEAN"; else echo "CODE FORMAT NOT CLEAN"; echo; echo "THE FOLLOWING FILES NEED TO BE FORMATTED"; echo; git ls-files -m; echo; if [[ $FDB_VERSION =~ 7\.\3. ]]; then echo skip; else exit 1; fi; fi. Reason: exit status 1
Build Log terminal output (available for 30 days)
Build Workspace zip file of the working directory (available for 30 days)

foundationdb-ci · 2026-05-27T17:02:53Z

Result of foundationdb-pr-clang on Linux RHEL 9

Commit ID: 9b4b2f5
Duration 0:04:20
Result: ❌ FAILED
Error: Error while executing command: if [[ $(git diff --shortstat 2> /dev/null | tail -n1) == "" ]]; then echo "CODE FORMAT CLEAN"; else echo "CODE FORMAT NOT CLEAN"; echo; echo "THE FOLLOWING FILES NEED TO BE FORMATTED"; echo; git ls-files -m; echo; if [[ $FDB_VERSION =~ 7\.\3. ]]; then echo skip; else exit 1; fi; fi. Reason: exit status 1
Build Log terminal output (available for 30 days)
Build Workspace zip file of the working directory (available for 30 days)

foundationdb-ci · 2026-05-27T17:02:58Z

Result of foundationdb-pr-cluster-tests on Linux RHEL 9

Commit ID: 9b4b2f5
Duration 0:04:28
Result: ❌ FAILED
Error: Error while executing command: if [[ $(git diff --shortstat 2> /dev/null | tail -n1) == "" ]]; then echo "CODE FORMAT CLEAN"; else echo "CODE FORMAT NOT CLEAN"; echo; echo "THE FOLLOWING FILES NEED TO BE FORMATTED"; echo; git ls-files -m; echo; if [[ $FDB_VERSION =~ 7\.\3. ]]; then echo skip; else exit 1; fi; fi. Reason: exit status 1
Build Log terminal output (available for 30 days)
Build Workspace zip file of the working directory (available for 30 days)
Cluster Test Logs zip file of the test logs (available for 30 days)

foundationdb-ci · 2026-05-27T17:02:58Z

Result of foundationdb-pr-clang-ide on Linux RHEL 9

Commit ID: 9b4b2f5
Duration 0:04:24
Result: ❌ FAILED
Error: Error while executing command: if [[ $(git diff --shortstat 2> /dev/null | tail -n1) == "" ]]; then echo "CODE FORMAT CLEAN"; else echo "CODE FORMAT NOT CLEAN"; echo; echo "THE FOLLOWING FILES NEED TO BE FORMATTED"; echo; git ls-files -m; echo; if [[ $FDB_VERSION =~ 7\.\3. ]]; then echo skip; else exit 1; fi; fi. Reason: exit status 1
Build Log terminal output (available for 30 days)
Build Workspace zip file of the working directory (available for 30 days)

…ive-fdb-cdc

foundationdb-ci · 2026-06-23T22:51:31Z

Result of foundationdb-pr-clang-ide on Linux RHEL 9

Commit ID: f9faf26
Duration 0:21:50
Result: ✅ SUCCEEDED
Error: N/A
Build Log terminal output (available for 30 days)
Build Workspace zip file of the working directory (available for 30 days)

tclinkenbeard-oai

What is it trying to do?

This adds native CDC support: durable named CDC streams over registered key ranges, CDC pseudo-tag routing in the commit path, CDC proxies that peek/filter/buffer/serve/ack/pop TLog data, shared-tag safe-pop semantics, recovery/failover support, and simulation coverage. Bindings are intentionally deferred.

Is it correct?

By inspection, yes. I traced the registration/removal metadata path, commit-path CDC tag injection, clear-range clipping, proxy buffering and delivery frontiers, durable acknowledgement reconciliation, active/retired tag popping, recovery retention, proxy recruitment, and primary/remote/satellite TLog propagation.

I also checked the earlier AI-review findings against the current head. The previously reported high/medium issues appear fixed or explicitly documented as intended behavior. I did not find a concrete correctness regression.

Are there bugs?

No confirmed bugs found.

Are there omissions?

No blocking omissions. The test coverage appears broad by inspection: lifecycle behavior, clear clipping, proxy replacement, memory bounds, durable acknowledgement scanning, retired-tag cleanup/recovery, disabled restart/drain, and satellite configurations.

I did not run builds, tests, simulations, or other validation commands.

Are there better ways of doing things?

The remaining notable limitation is scalability: registration, safe-pop reconstruction, and assignment publication still scan global CDC metadata in single transactions. The design already identifies this as a scaling boundary. A future follow-up could maintain per-tag aggregate safe-pop state and an indexed assignment structure so these paths become incremental.

Should this CL be LGTMd?

Yes, subject to the pending public CI completing successfully.

Current public CI has no failures: clang-format and Test Boost CONFIG Mode on Windows pass; the default, clang, clang-arm, clang-ide, macOS, macOS_m1, and cluster-test builders are still pending.

foundationdb-ci · 2026-06-23T23:03:10Z

Result of foundationdb-pr-macos-m1 on macOS 14.x

Commit ID: f9faf26
Duration 0:33:28
Result: ✅ SUCCEEDED
Error: N/A
Build Log terminal output (available for 30 days)
Build Workspace zip file of the working directory (available for 30 days)

tclinkenbeard-oai · 2026-06-23T23:03:32Z

@gxglass I mostly added new test code since the last review, but re-requesting review now while I run another filtered 100k correctness test ensemble, because the PR has grown more

foundationdb-ci · 2026-06-23T23:15:34Z

Result of foundationdb-pr-clang-arm on Linux RHEL 9

Commit ID: f9faf26
Duration 0:45:52
Result: ✅ SUCCEEDED
Error: N/A
Build Log terminal output (available for 30 days)
Build Workspace zip file of the working directory (available for 30 days)

foundationdb-ci · 2026-06-23T23:18:31Z

Result of foundationdb-pr on Linux RHEL 9

Commit ID: f9faf26
Duration 0:48:47
Result: ❌ FAILED
Error: Error while executing command: if python3 -m joshua.joshua list --stopped | grep ${ENSEMBLE_ID} | grep -q 'pass=10[0-9][0-9][0-9]'; then echo PASS; else echo FAIL && exit 1; fi. Reason: exit status 1
Build Log terminal output (available for 30 days)
Build Workspace zip file of the working directory (available for 30 days)

foundationdb-ci · 2026-06-23T23:25:33Z

Result of foundationdb-pr-clang on Linux RHEL 9

Commit ID: f9faf26
Duration 0:55:51
Result: ❌ FAILED
Error: Error while executing command: if python3 -m joshua.joshua list --stopped | grep ${ENSEMBLE_ID} | grep -q 'pass=10[0-9][0-9][0-9]'; then echo PASS; else echo FAIL && exit 1; fi. Reason: exit status 1
Build Log terminal output (available for 30 days)
Build Workspace zip file of the working directory (available for 30 days)

foundationdb-ci · 2026-06-23T23:33:20Z

Result of foundationdb-pr-cluster-tests on Linux RHEL 9

Commit ID: f9faf26
Duration 1:03:40
Result: ✅ SUCCEEDED
Error: N/A
Build Log terminal output (available for 30 days)
Build Workspace zip file of the working directory (available for 30 days)
Cluster Test Logs zip file of the test logs (available for 30 days)

gxglass · 2026-06-23T23:42:56Z

Here's the current review from local AI. This diff: 60 files changed, 7529 insertions(+), 91 deletions(-).
Human review still pending (took a bit of a delay this week)

pr13287-cdc-review-round3.md

foundationdb-ci · 2026-06-24T02:36:44Z

Result of foundationdb-pr-macos on macOS 14.x

Commit ID: f9faf26
Duration 4:07:02
Result: ✅ SUCCEEDED
Error: N/A
Build Log terminal output (available for 30 days)
Build Workspace zip file of the working directory (available for 30 days)

foundationdb-ci · 2026-06-24T05:07:27Z

Result of foundationdb-pr-clang-ide on Linux RHEL 9

Commit ID: eb75790
Duration 0:22:04
Result: ✅ SUCCEEDED
Error: N/A
Build Log terminal output (available for 30 days)
Build Workspace zip file of the working directory (available for 30 days)

foundationdb-ci · 2026-06-24T05:18:41Z

Result of foundationdb-pr-macos-m1 on macOS 14.x

Commit ID: eb75790
Duration 0:33:18
Result: ✅ SUCCEEDED
Error: N/A
Build Log terminal output (available for 30 days)
Build Workspace zip file of the working directory (available for 30 days)

foundationdb-ci · 2026-06-24T05:33:08Z

Result of foundationdb-pr-clang-arm on Linux RHEL 9

Commit ID: eb75790
Duration 0:47:44
Result: ✅ SUCCEEDED
Error: N/A
Build Log terminal output (available for 30 days)
Build Workspace zip file of the working directory (available for 30 days)

foundationdb-ci · 2026-06-24T05:33:11Z

Result of foundationdb-pr-macos on macOS 14.x

Commit ID: eb75790
Duration 0:47:50
Result: ✅ SUCCEEDED
Error: N/A
Build Log terminal output (available for 30 days)
Build Workspace zip file of the working directory (available for 30 days)

foundationdb-ci · 2026-06-24T05:44:53Z

Result of foundationdb-pr-clang on Linux RHEL 9

Commit ID: eb75790
Duration 0:59:26
Result: ✅ SUCCEEDED
Error: N/A
Build Log terminal output (available for 30 days)
Build Workspace zip file of the working directory (available for 30 days)

foundationdb-ci · 2026-06-24T05:46:37Z

Result of foundationdb-pr on Linux RHEL 9

Commit ID: eb75790
Duration 1:01:16
Result: ✅ SUCCEEDED
Error: N/A
Build Log terminal output (available for 30 days)
Build Workspace zip file of the working directory (available for 30 days)

foundationdb-ci · 2026-06-24T05:48:35Z

Result of foundationdb-pr-cluster-tests on Linux RHEL 9

Commit ID: eb75790
Duration 1:03:11
Result: ✅ SUCCEEDED
Error: N/A
Build Log terminal output (available for 30 days)
Build Workspace zip file of the working directory (available for 30 days)
Cluster Test Logs zip file of the test logs (available for 30 days)

gxglass

Partial review comments follow. This round was essentially just a close second reading of the design doc with a couple Ctrl-F searches to discover knob values for clarity.

This is obviously a very good design doc. It describes the system from all important angles at the right level of detail to convince a reader that the choices are well motivated and can be pulled off. The design itself seems sound and fits in well with how FDB works.

Comments below reflect a few clarity nits and future-work-documentation suggestions, and a question or two. None of this gates LGTM or points to risks/gaps/etc. It's basically just doc enhancement related.

gxglass · 2026-06-25T23:11:33Z

+`fdbclient/CDCProxyInterface.h`.
+
+```cpp
+Future<CDCStreamId> registerNativeCdcStreamClient(Database cx, Key name, KeyRange keys);


It helps here and below if readers know that CDCStreamId is a uint64_t typedef (I had to look it up)

gxglass · 2026-06-25T23:48:19Z

+per-stream in-memory buffer.
+
+All raw peek windows and stream buffers owned by one CDC proxy share a
+`CDC_PROXY_BUFFER_BYTES` budget. A replicated log read may retain one separately


mention production default 1GB

gxglass · 2026-06-25T23:58:24Z

+## Acknowledgement and tag popping
+
+Acknowledgement is per stream, while TLog popping is per CDC tag. This
+distinction is the core retention rule.


Nice

also: it's fifos all the way down

gxglass · 2026-06-26T00:01:02Z

+retention decision.
+
+Acknowledgement notifications are level-triggered and coalesced for at least
+`CDC_PROXY_POP_MIN_INTERVAL`. A notification received during a durable-state


mention 100ms default for clarity

gxglass · 2026-06-26T00:14:26Z

+3. For a new name, it validates the feature knob.
+4. It allocates a new monotonically increasing `CDCStreamId`.
+5. It selects a CDC tag using current active stream counts. The allocator uses
+   the least populated tag among `NATIVE_CDC_TAG_COUNT` tags, choosing the


give production default (256) here

gxglass · 2026-06-26T00:15:14Z

+differs.
+
+`NATIVE_CDC_TAG_COUNT` controls the bounded tag pool used for new stream
+allocation and must be between 1 and 65,536 inclusive. Invalid values reject


Prefer to discuss considerations about how a good default can be chosen. (Also it's totally unclear why somebody would want 64K tags.) I'm guessing the default is chosen by something like the following: it's roughly 10x more than the number of proxies we anticipate, so that load balancing of tags to processes ends up with proxies being roughly evenly balanced even if moderate (small mulitple) per-tag variance in write traffic exists.

gxglass · 2026-06-26T00:28:37Z

+  proxy throughput, buffer memory, lag, or number of active readers.
+* Assignment mutations use one coalescing change key that wakes a full durable
+  ownership rescan. This is appropriate for low-rate control-plane changes but
+  should be sharded if registration and removal throughput becomes material.


API users should perhaps be aware of this -- maybe mention expected registrations/sec in the API example section much earlier

gxglass · 2026-06-26T00:30:09Z

+  response to load. A future implementation can use versioned tag history to
+  make such changes without losing the ability to read earlier tagged data.
+* The native interface does not yet provide external binding support,
+  administrative tooling, or a higher-level consumer checkpoint abstraction.


Is there a command to force unregister a stream? If not maybe a short section on "Needed for production" stuff (which I think this would be?)

gxglass · 2026-06-26T00:36:38Z

+abandoned consumer, but it would silently violate the stated retention
+contract for a slow active stream. The initial design therefore requires an
+acknowledgement or explicit removal before releasing required history and
+treats administrative expiration policy as future work.


Somehow I'm guessing this is going to need to be invented before long, but of course it's fine to pipeline the work.

Scenario: CDC downstream/consumer system is having trouble, AND FDB performance under tlog spilling is borderline overloaded for a cluster in question. What do we want to happen? I'm guessing some manner of "turn off CDC and recompute downstream application state from a full scan of the database" once the downstream system gets generally unwedged. At least some of the time this would probably be desired. That said, maybe I'm wrong about TLOG performance / spilling being considered something to generally avoid. My impression is we try pretty hard to avoid that locally.

gxglass · 2026-06-26T00:38:49Z

+acknowledgement lag, safe-pop distance, CDC-attributed TLog retention, and proxy
+buffer pressure. The oldest required stream ID identifies the first consumer
+to investigate. Because the initial implementation has no automatic stream
+expiration, operators must repair the consumer, explicitly discard its


Mentioned above but some table of anticipated tooling to guide future development work may be helpful here.

gxglass

Review through fdbclient/.

The natural diff ordering here represents a reasonable review order: design/, fdbclient/, fdbserver/, fdbserver/workloads, tests/, so I'll just go in that order.

NOTE: a couple comments below say "uint64_t" when they mean to say "uint16_t".

gxglass · 2026-06-26T18:12:57Z

+#include "fdbclient/CommitTransaction.h"
+#include "flow/FileIdentifier.h"
+#include "fdbrpc/fdbrpc.h"
+


minor: suggest a link to design/cdc.md re: this protocol & its messages

gxglass · 2026-06-26T18:18:06Z

 Optional<Value> decodeTagLocalityListKey(KeyRef const&);
 int8_t decodeTagLocalityListValue(ValueRef const&);

+// Native CDC stream routing and lifecycle metadata persisted in the transaction state store.


this stuff is well described in the design doc so a comment link here too could be useful

gxglass · 2026-06-26T18:52:51Z

+
+bool validNativeCdcTagCount(int tagCount) {
+	return tagCount > 0 &&
+	       static_cast<uint64_t>(tagCount) <= static_cast<uint64_t>(std::numeric_limits<uint16_t>::max()) + 1;


suggest a typedef for the uint64_t because it's used below and drift here would be bad

gxglass · 2026-06-26T18:53:26Z

+class NativeCdcIdentifierAllocator {
+	bool sawStream = false;
+	CDCStreamId maxStreamId = 0;
+	std::unordered_map<uint16_t, uint32_t> tagStreamCounts;


suggest uint64_t ==> CDCTagId_t or something. also increases clarity a little

gxglass · 2026-06-26T19:00:08Z

+}
+
+void signalNativeCdcProxyAssignmentChange(Transaction* tr) {
+	// Assignment updates are low-rate control-plane operations. A single


could put a metric to count this to be able to confirm it's actually low rate

gxglass · 2026-06-26T19:08:52Z

+			const Key nameKey = cdcStreamNameKeyFor(name);
+			Optional<Value> currentId = co_await tr.get(nameKey);
+			if (!nativeCdcNameMatchesStream(currentId, streamId)) {
+				CODE_PROBE(currentId.present(), "Native CDC preserves a replacement stream during removal retry");


I don't know how rare this is expected to be but logging something here may be useful

foundationdb-ci · 2026-06-26T21:22:16Z

Result of foundationdb-pr-clang-arm on Linux RHEL 9

Commit ID: b1e3e87
Duration 0:46:01
Result: ✅ SUCCEEDED
Error: N/A
Build Log terminal output (available for 30 days)
Build Workspace zip file of the working directory (available for 30 days)

foundationdb-ci · 2026-06-26T21:22:58Z

Result of foundationdb-pr-clang-ide on Linux RHEL 9

Commit ID: b1e3e87
Duration 0:46:45
Result: ✅ SUCCEEDED
Error: N/A
Build Log terminal output (available for 30 days)
Build Workspace zip file of the working directory (available for 30 days)

foundationdb-ci · 2026-06-26T21:47:23Z

Result of foundationdb-pr on Linux RHEL 9

Commit ID: b1e3e87
Duration 1:11:09
Result: ✅ SUCCEEDED
Error: N/A
Build Log terminal output (available for 30 days)
Build Workspace zip file of the working directory (available for 30 days)

foundationdb-ci · 2026-06-26T21:51:33Z

Result of foundationdb-pr-cluster-tests on Linux RHEL 9

Commit ID: b1e3e87
Duration 1:15:18
Result: ✅ SUCCEEDED
Error: N/A
Build Log terminal output (available for 30 days)
Build Workspace zip file of the working directory (available for 30 days)
Cluster Test Logs zip file of the test logs (available for 30 days)

foundationdb-ci · 2026-06-26T22:04:45Z

Result of foundationdb-pr-clang on Linux RHEL 9

Commit ID: b1e3e87
Duration 1:28:31
Result: ✅ SUCCEEDED
Error: N/A
Build Log terminal output (available for 30 days)
Build Workspace zip file of the working directory (available for 30 days)

foundationdb-ci · 2026-06-26T22:16:56Z

Result of foundationdb-pr-macos-m1 on macOS 14.x

Commit ID: b1e3e87
Duration 1:40:44
Result: ✅ SUCCEEDED
Error: N/A
Build Log terminal output (available for 30 days)
Build Workspace zip file of the working directory (available for 30 days)

foundationdb-ci · 2026-06-27T00:48:35Z

Result of foundationdb-pr-macos on macOS 14.x

Commit ID: b1e3e87
Duration 4:12:19
Result: ❌ FAILED
Error: `Error while executing command: # compatible with bash and zsh
ssh -o StrictHostKeyChecking=no -o UserKnownHostsFile=/dev/null -i ${HOME}/.ssh_key
-o ServerAliveInterval=9 ec2-user@${MAC_EC2_HOST} "
set -e -o noclobber
BUILDNUM=${CODEBUILD_BUILD_NUMBER}
WAIT_TIME=14400 # 4 hours
LOCK_STALE=4500 # 75 minutes
LOCK=/tmp/ci.lock
"'

macOS stat flags, not linux compatible

lock_mtime() { stat -f %m $LOCK || echo 0 }
I=0
while ! (echo $BUILDNUM >$LOCK) 2>/dev/null ; do
if (( ++I >= WAIT_TIME )); then
echo "timeout waiting for ci.lock after $I seconds"
exit 1
fi
if (( $(date +%s) - $(lock_mtime) >= LOCK_STALE )); then
echo "lock expired after $LOCK_STALE seconds ..."
# racy but sleep so chances are low
sleep $(( 3 + RANDOM % 8 ))
if (( $(date +%s) - $(lock_mtime) >= LOCK_STALE )); then
echo "lock holder is gone"
rm -f $LOCK
else
echo "new lock holder"
fi
fi
sleep 1
done
echo "acquired ci.lock"
'
. Reason: exit status 1`

Build Log terminal output (available for 30 days)
Build Workspace zip file of the working directory (available for 30 days)

tclinkenbeard-oai added 26 commits May 27, 2026 05:57

Initial native CDC commit

d86ffe0

Add native CDC commit proxy routing

158333e

Add native CDC stream lifecycle metadata operations and tests

9ce0954

Fix correctness regressions

a28711d

Add native CDC acknowledgement watermarks

46e41a0

Add native CDC proxy consume and acknowledge backend

0bc5799

Recruit and publish CDC proxy endpoints through recovery routing

7203fe5

Persist and publish CDC stream ownership routing

9f52b0c

Add non-recovery CDC proxy failover and durable stream reassignment

9b76882

Preserve native CDC streams across recovery and proxy failover

230af4d

Add buffered native CDC consumption and safe shared-tag acknowledgements

92752c5

Add client-facing native CDC operations with proxy failover retries

4d48b1c

Finalize CDC retention state when streams are removed

e61abe0

Bound buffered mutation memory in CDC proxies

364ec1e

Avoid waiting forever during concurrent CDC stream removal

a8599a4

Bound CDC tags and persist stream ID allocation

1808150

Fix CDC proxy buffered byte release accounting

7acc707

Resume CDC buffers from the retained minimum version

dfb79ba

Refresh CDC buffer cursors after acknowledgements

f98c7f5

Anchor CDC retention at stream registration commit versions

7ada95e

Register the CDC shared-tag simulator test

d3c942d

Fix Native CDC recovery progress across log generations

eb58b1a

Contain Native CDC stream expiry without terminating its proxy

7a9e54e

Fix Native CDC workload consumption synchronization

f26afd0

Fix Native CDC recovery handling and stabilize simulator coverage

e0d06a4

Simplify serialize methods

9b4b2f5

Merge remote-tracking branch 'origin/main' into dev/tclinkenbeard/nat…

f9faf26

…ive-fdb-cdc

tclinkenbeard-oai commented Jun 23, 2026

View reviewed changes

tclinkenbeard-oai requested a review from gxglass June 23, 2026 23:03

Remove unstable Native CDC recovery test hook

eb75790

gxglass reviewed Jun 26, 2026

View reviewed changes

Address native CDC review follow-ups

b1e3e87

Uh oh!

Conversation

tclinkenbeard-oai commented May 27, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

foundationdb-ci commented May 27, 2026

Result of foundationdb-pr-clang-arm on Linux CentOS 7

Uh oh!

foundationdb-ci commented May 27, 2026

Result of foundationdb-pr-clang on Linux RHEL 9

Uh oh!

foundationdb-ci commented May 27, 2026

Result of foundationdb-pr-cluster-tests on Linux RHEL 9

Uh oh!

foundationdb-ci commented May 27, 2026

Result of foundationdb-pr-clang-ide on Linux RHEL 9

Uh oh!

foundationdb-ci commented Jun 23, 2026

Result of foundationdb-pr-clang-ide on Linux RHEL 9

Uh oh!

tclinkenbeard-oai left a comment

Choose a reason for hiding this comment

What is it trying to do?

Is it correct?

Are there bugs?

Are there omissions?

Are there better ways of doing things?

Should this CL be LGTMd?

Uh oh!

foundationdb-ci commented Jun 23, 2026

Result of foundationdb-pr-macos-m1 on macOS 14.x

Uh oh!

tclinkenbeard-oai commented Jun 23, 2026

Uh oh!

foundationdb-ci commented Jun 23, 2026

Result of foundationdb-pr-clang-arm on Linux RHEL 9

Uh oh!

foundationdb-ci commented Jun 23, 2026

Result of foundationdb-pr on Linux RHEL 9

Uh oh!

foundationdb-ci commented Jun 23, 2026

Result of foundationdb-pr-clang on Linux RHEL 9

Uh oh!

foundationdb-ci commented Jun 23, 2026

Result of foundationdb-pr-cluster-tests on Linux RHEL 9

Uh oh!

gxglass commented Jun 23, 2026

Uh oh!

foundationdb-ci commented Jun 24, 2026

Result of foundationdb-pr-macos on macOS 14.x

Uh oh!

foundationdb-ci commented Jun 24, 2026

Result of foundationdb-pr-clang-ide on Linux RHEL 9

Uh oh!

foundationdb-ci commented Jun 24, 2026

Result of foundationdb-pr-macos-m1 on macOS 14.x

Uh oh!

foundationdb-ci commented Jun 24, 2026

Result of foundationdb-pr-clang-arm on Linux RHEL 9

Uh oh!

foundationdb-ci commented Jun 24, 2026

Result of foundationdb-pr-macos on macOS 14.x

Uh oh!

foundationdb-ci commented Jun 24, 2026

Result of foundationdb-pr-clang on Linux RHEL 9

Uh oh!

foundationdb-ci commented Jun 24, 2026

Result of foundationdb-pr on Linux RHEL 9

Uh oh!

foundationdb-ci commented Jun 24, 2026

Result of foundationdb-pr-cluster-tests on Linux RHEL 9

Uh oh!

gxglass left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

tclinkenbeard-oai commented May 27, 2026 •

edited

Loading