Sudden Codebase Indexing issue - Handshake failed

Discussion
1

Hello lovely Cursor Community,

I know there are sometimes bugs and Cursor isn’t perfect, but most of the time it functions well. What I’m experiencing right now is that the codebase indexing function suddenly stopped working out of nowhere. Yesterday, I worked on a couple of projects, and it worked perfectly. However, since today, it seems that it doesn’t work anymore. I always receive the message “Handshake failed.”

Is anyone else experiencing the same issue? I’ve read through all the articles from previous users who mentioned this, but it seems like no one has found a solution.

Any help would be appreciated.

Indexing fail
Indexing fail976×515 16 KB

8 Likes
2

Yep dosent work.

2024-05-25 14:32:14.675 [info] Handshake start
2024-05-25 14:32:22.939 [info] Handshake start
2024-05-25 14:32:39.210 [info] Handshake start
2024-05-25 14:33:11.309 [info] Handshake start
2024-05-25 14:34:15.577 [error] Handshake failed:
2024-05-25 14:34:15.577 [error] Error: timeout in handshake with retry
at e.RepoClientMultiplexer.handshakeWithRetry (/Applications/Cursor.app/Contents/Resources/app/extensions/cursor-retrieval/dist/main.js:2:995054)
at async m.getServerStatus (/Applications/Cursor.app/Contents/Resources/app/extensions/cursor-retrieval/dist/main.js:2:1009958)
at async m.startIndexingRepository (/Applications/Cursor.app/Contents/Resources/app/extensions/cursor-retrieval/dist/main.js:2:1012203)
2024-05-25 14:34:15.577 [error] Handshake failed.
2024-05-25 14:34:15.578 [info] Indexing job successfully done or aborted.

1 Like
3

I’m having the same issue right now :d

4

Just confirming the same on a new install of Cursor has the same issue. Upgraded to Pro to save on my own API costs and did not affect the indexing (not that I expected it to).

5

Small infra issue with indexing at the moment, will have it back in an our or so.

1 Like
6

Should be fixed! Let us know if you see anything off on your end.

7

Works now :+1:

8

Same issue…

9

same problem

10

Hey Cursor Team,

I am suddenly seeing the same issue now.

Error indexing codebase
Codebase chat is falling back to BM25, which is slower and less accurate than embeddings.
2 Likes
11

did you manage to resolve it? I have the same issue

12

I am getting the same issue

13

I have got same issue today. Tried deleting index, and no use.
But I have got very big conversation in one chat. Maybe I need to start a new chat, maybe because of the context is too big for current chat.

14

I’m having the same indexing handshake issue. Up until the last update everything was working perfectly.

After the last update (sometime within this last week I believe) I get handshake errors on two separate Windows 10 machines on two different networks. The issue seems to be recent and constant.

I’ve tried disabling HTTP2, deleted indexes, restarted cursor, reboot machines. Nothing is correcting the indexing errors.

My version info.
Version: 0.41.3
VSCode Version: 1.91.1
Commit: 51c8aff7cb5a89f4a0e462fbacab938bdbfaf140
Date: 2024-09-25T17:13:50.377Z
Electron: 29.4.0
ElectronBuildId: undefined
Chromium: 122.0.6261.156
Node.js: 20.9.0
V8: 12.2.281.27-electron.0
OS: Windows_NT x64 10.0.19045

Pro account.

15

I am getting the same error too. What is the issue. Is it outage or I have to do something on my system to build this index.
Reinstall cursor did not helped me.
Pro account

Error indexing codebase

Open Settings

Codebase chat is falling back to BM25, which is slower and less accurate than embeddings.

1 Like
16

image
image1030×450 18.5 KB

1 Like
17

Could be outage on their server. not 100% sure about this issue though

18

I shell be very thankful If you can help me what resolved this issue.
I am facing same issue.

1 Like
19

I’m also having the same issue.

App details:
Version: 0.41.3
VSCode Version: 1.91.1
Commit: 51c8aff7cb5a89f4a0e462fbacab938bdbfaf140
Date: 2024-09-25T17:13:50.377Z
Electron: 29.4.0
ElectronBuildId: undefined
Chromium: 122.0.6261.156
Node.js: 20.9.0
V8: 12.2.281.27-electron.0
OS: Darwin arm64 23.1.0

Output trace:
2024-10-08 13:05:22.522 [info] Creating merkle client.

2024-10-08 13:05:22.522 [info] Done creating merkle client.

2024-10-08 13:05:22.522 [info] Doing a startup handshake.

2024-10-08 13:05:22.673 [info] Finished initializing merkle tree in 151.75770902633667 ms.

2024-10-08 13:05:22.674 [info] Doing the initial handshake with hash: 4f56d56b846061c27aa529ee4b399a6b909d6195452901ead203311dfc7ea170

2024-10-08 13:05:22.674 [info] Handshake start

2024-10-08 13:05:22.953 [info] Handshake timing: 279.2430419921875

2024-10-08 13:05:24.957 [info] Handshake start

2024-10-08 13:05:25.208 [info] Handshake timing: 251.09337496757507

2024-10-08 13:05:29.208 [info] Handshake start

2024-10-08 13:05:29.470 [info] Handshake timing: 261.72062492370605

2024-10-08 13:05:37.475 [info] Handshake start

2024-10-08 13:05:37.745 [info] Handshake timing: 270.0830409526825

2024-10-08 13:05:53.750 [info] Handshake start

2024-10-08 13:05:53.997 [info] Handshake timing: 246.9032919406891

2024-10-08 13:06:26.001 [info] Handshake start

2024-10-08 13:06:26.260 [info] Handshake timing: 259.888375043869

2024-10-08 13:07:30.263 [error] Handshake failed:

2024-10-08 13:07:30.263 [error] Error: timeout in handshake with retry

at t.RepoClientMultiplexer.handshakeWithRetry (/Applications/Cursor.app/Contents/Resources/app/extensions/cursor-retrieval/dist/main.js:2:826553)

at async w.getServerStatus (/Applications/Cursor.app/Contents/Resources/app/extensions/cursor-retrieval/dist/main.js:2:841881)

at async w.startIndexingRepository (/Applications/Cursor.app/Contents/Resources/app/extensions/cursor-retrieval/dist/main.js:2:844181)

2024-10-08 13:07:30.263 [error] Handshake failed.

2024-10-08 13:07:30.263 [info] Indexing job successfully done or aborted.

4 Likes
20

@amanrs is this an issue again? I’ve had this happen multiple times today.

2 Likes