👥 fix(assistants): Improve Error handling (#2012)

* feat: make assistants endpoint appendable since message state is not managed by LibreChat * fix(ask): search currentMessages for thread_id if it's not defined * refactor(abortMiddleware): remove use of `overrideProps` and spread unknown fields instead * chore: remove console.log in `abortConversation` * refactor(assistants): improve error handling/cancellation flow
danny-avila · Mar 7, 2024 · 18edd26 · 18edd26
1 parent d4fe8fc
commit 18edd26
Show file tree

Hide file tree

Showing 7 changed files with 77 additions and 25 deletions.
diff --git a/api/server/middleware/abortMiddleware.js b/api/server/middleware/abortMiddleware.js
@@ -110,7 +110,7 @@ const handleAbortError = async (res, req, error, data) => {
   }
 
   const respondWithError = async (partialText) => {
-    const options = {
+    let options = {
       sender,
       messageId,
       conversationId,
@@ -121,7 +121,8 @@ const handleAbortError = async (res, req, error, data) => {
     };
 
     if (partialText) {
-      options.overrideProps = {
+      options = {
+        ...options,
         error: false,
         unfinished: true,
         text: partialText,

diff --git a/api/server/middleware/abortRun.js b/api/server/middleware/abortRun.js
@@ -37,7 +37,7 @@ async function abortRun(req, res) {
   try {
     await cache.set(cacheKey, 'cancelled');
     const cancelledRun = await openai.beta.threads.runs.cancel(thread_id, run_id);
-    logger.debug('Cancelled run:', cancelledRun);
+    logger.debug('[abortRun] Cancelled run:', cancelledRun);
   } catch (error) {
     logger.error('[abortRun] Error cancelling run', error);
     if (

diff --git a/api/server/routes/assistants/chat.js b/api/server/routes/assistants/chat.js
@@ -11,10 +11,10 @@ const {
 } = require('~/server/services/Threads');
 const { runAssistant, createOnTextProgress } = require('~/server/services/AssistantService');
 const { addTitle, initializeClient } = require('~/server/services/Endpoints/assistant');
+const { sendResponse, sendMessage } = require('~/server/utils');
 const { createRun, sleep } = require('~/server/services/Runs');
 const { getConvo } = require('~/models/Conversation');
 const getLogStores = require('~/cache/getLogStores');
-const { sendMessage } = require('~/server/utils');
 const { logger } = require('~/config');
 
 const router = express.Router();
@@ -101,32 +101,52 @@ router.post('/', validateModel, buildEndpointOption, setHeaders, async (req, res
   let completedRun;
 
   const handleError = async (error) => {
+    const messageData = {
+      thread_id,
+      assistant_id,
+      conversationId,
+      parentMessageId,
+      sender: 'System',
+      user: req.user.id,
+      shouldSaveMessage: false,
+      messageId: responseMessageId,
+      endpoint: EModelEndpoint.assistants,
+    };
+
     if (error.message === 'Run cancelled') {
       return res.end();
-    }
-    if (error.message === 'Request closed' && completedRun) {
+    } else if (error.message === 'Request closed' && completedRun) {
       return;
     } else if (error.message === 'Request closed') {
       logger.debug('[/assistants/chat/] Request aborted on close');
+    } else {
+      logger.error('[/assistants/chat/]', error);
     }
 
-    logger.error('[/assistants/chat/]', error);
-
     if (!openai || !thread_id || !run_id) {
-      return res.status(500).json({ error: 'The Assistant run failed to initialize' });
+      return sendResponse(res, messageData, 'The Assistant run failed to initialize');
     }
 
+    await sleep(3000);
+
     try {
+      const status = await cache.get(cacheKey);
+      if (status === 'cancelled') {
+        logger.debug('[/assistants/chat/] Run already cancelled');
+        return res.end();
+      }
       await cache.delete(cacheKey);
       const cancelledRun = await openai.beta.threads.runs.cancel(thread_id, run_id);
-      logger.debug('Cancelled run:', cancelledRun);
+      logger.debug('[/assistants/chat/] Cancelled run:', cancelledRun);
     } catch (error) {
-      logger.error('[abortRun] Error cancelling run', error);
+      logger.error('[/assistants/chat/] Error cancelling run', error);
     }
 
     await sleep(2000);
+
+    let run;
     try {
-      const run = await openai.beta.threads.runs.retrieve(thread_id, run_id);
+      run = await openai.beta.threads.runs.retrieve(thread_id, run_id);
       await recordUsage({
         ...run.usage,
         model: run.model,
@@ -137,6 +157,7 @@ router.post('/', validateModel, buildEndpointOption, setHeaders, async (req, res
       logger.error('[/assistants/chat/] Error fetching or processing run', error);
     }
 
+    let finalEvent;
     try {
       const runMessages = await checkMessageGaps({
         openai,
@@ -146,22 +167,18 @@ router.post('/', validateModel, buildEndpointOption, setHeaders, async (req, res
         latestMessageId: responseMessageId,
       });
 
-      const finalEvent = {
+      finalEvent = {
         title: 'New Chat',
         final: true,
         conversation: await getConvo(req.user.id, conversationId),
         runMessages,
       };
-
-      if (res.headersSent && finalEvent) {
-        return sendMessage(res, finalEvent);
-      }
-
-      res.json(finalEvent);
     } catch (error) {
       logger.error('[/assistants/chat/] Error finalizing error process', error);
-      return res.status(500).json({ error: 'The Assistant run failed' });
+      return sendResponse(res, messageData, 'The Assistant run failed');
     }
+
+    return sendResponse(res, finalEvent);
   };
 
   try {
@@ -172,10 +189,12 @@ router.post('/', validateModel, buildEndpointOption, setHeaders, async (req, res
     });
 
     if (convoId && !_thread_id) {
+      completedRun = true;
       throw new Error('Missing thread_id for existing conversation');
     }
 
     if (!assistant_id) {
+      completedRun = true;
       throw new Error('Missing assistant_id');
     }
 

diff --git a/api/server/utils/streamResponse.js b/api/server/utils/streamResponse.js
@@ -32,6 +32,13 @@ const sendMessage = (res, message, event = 'message') => {
  * @async
  * @param {object} res - The server response.
  * @param {object} options - The options for handling the error containing message properties.
+ * @param {object} options.user - The user ID.
+ * @param {string} options.sender - The sender of the message.
+ * @param {string} options.conversationId - The conversation ID.
+ * @param {string} options.messageId - The message ID.
+ * @param {string} options.parentMessageId - The parent message ID.
+ * @param {string} options.text - The error message.
+ * @param {boolean} options.shouldSaveMessage - [Optional] Whether the message should be saved. Default is true.
  * @param {function} callback - [Optional] The callback function to be executed.
  */
 const sendError = async (res, options, callback) => {
@@ -43,7 +50,7 @@ const sendError = async (res, options, callback) => {
     parentMessageId,
     text,
     shouldSaveMessage,
-    overrideProps = {},
+    ...rest
   } = options;
   const errorMessage = {
     sender,
@@ -55,7 +62,7 @@ const sendError = async (res, options, callback) => {
     final: true,
     text,
     isCreatedByUser: false,
-    ...overrideProps,
+    ...rest,
   };
   if (callback && typeof callback === 'function') {
     await callback();
@@ -88,7 +95,28 @@ const sendError = async (res, options, callback) => {
   handleError(res, errorMessage);
 };
 
+/**
+ * Sends the response based on whether headers have been sent or not.
+ * @param {Express.Response} res - The server response.
+ * @param {Object} data - The data to be sent.
+ * @param {string} [errorMessage] - The error message, if any.
+ */
+const sendResponse = (res, data, errorMessage) => {
+  if (!res.headersSent) {
+    if (errorMessage) {
+      return res.status(500).json({ error: errorMessage });
+    }
+    return res.json(data);
+  }
+
+  if (errorMessage) {
+    return sendError(res, { ...data, text: errorMessage });
+  }
+  return sendMessage(res, data);
+};
+
 module.exports = {
+  sendResponse,
   handleError,
   sendMessage,
   sendError,

diff --git a/client/src/hooks/Input/useTextarea.ts b/client/src/hooks/Input/useTextarea.ts
@@ -45,7 +45,9 @@ export default function useTextarea({
   const localize = useLocalize();
 
   const { conversationId, jailbreak, endpoint = '', assistant_id } = conversation || {};
-  const isNotAppendable = (latestMessage?.unfinished && !isSubmitting) || latestMessage?.error;
+  const isNotAppendable =
+    ((latestMessage?.unfinished && !isSubmitting) || latestMessage?.error) &&
+    endpoint !== EModelEndpoint.assistants;
   // && (conversationId?.length ?? 0) > 6; // also ensures that we don't show the wrong placeholder
 
   const assistant = endpoint === EModelEndpoint.assistants && assistantMap?.[assistant_id ?? ''];

diff --git a/client/src/hooks/SSE/useSSE.ts b/client/src/hooks/SSE/useSSE.ts
@@ -418,7 +418,6 @@ export default function useSSE(submission: TSubmission | null, index = 0) {
 
   const abortConversation = useCallback(
     async (conversationId = '', submission: TSubmission) => {
-      console.log(submission);
       let runAbortKey = '';
       try {
         const conversation = (JSON.parse(localStorage.getItem('lastConversationSetup') ?? '') ??

diff --git a/client/src/hooks/useChatHelpers.ts b/client/src/hooks/useChatHelpers.ts
@@ -140,7 +140,10 @@ export default function useChatHelpers(index = 0, paramId: string | undefined) {
       (msg) => msg.messageId === latestMessage?.parentMessageId,
     );
 
-    const thread_id = parentMessage?.thread_id ?? latestMessage?.thread_id;
+    let thread_id = parentMessage?.thread_id ?? latestMessage?.thread_id;
+    if (!thread_id) {
+      thread_id = currentMessages.find((message) => message.thread_id)?.thread_id;
+    }
 
     const endpointsConfig = queryClient.getQueryData<TEndpointsConfig>([QueryKeys.endpoints]);
     const endpointType = getEndpointField(endpointsConfig, endpoint, 'type');