Agents get several error only to chat with gpt4 o
dels] array (
'message' => 'Can\'t add messages to thread_miAYLuv6mPkVt6kpPItOEPNA while a run run_FdO49bJk2wvtPTESUSjWGnWG is active.',
'type' => 'invalid_request_error',
'param' => NULL,
'code' => NULL,
)
Sep 16 02:17:19 [26] /cmts.php :
[Models] array (
'message' => 'Thread thread_miAYLuv6mPkVt6kpPItOEPNA already has an active run run_FdO49bJk2wvtPTESUSjWGnWG.',
'type' => 'invalid_request_error',
'param' => NULL,
'code' => NULL,
)
Sep 16 02:17:19 [26] /cmts.php :
[Models] array (
'message' => 'Can\'t add messages to thread_miAYLuv6mPkVt6kpPItOEPNA while a run run_FdO49bJk2wvtPTESUSjWGnWG is active.',
'type' => 'invalid_request_error',
'param' => NULL,
'code' => NULL,
)
Sep 16 02:17:20 [26] /cmts.php :
[Models] array (
'message' => 'Thread thread_miAYLuv6mPkVt6kpPItOEPNA already has an active run run_FdO49bJk2wvtPTESUSjWGnWG.',
'type' => 'invalid_request_error',
'param' => NULL,
'code' => NULL,
)
this error led to a database crash after a short discussion with the agent in the backend studio, and there is something else, the agent posts my discussions on the public timeline, this seems absurd to me
-
-
·
LeonidS
- ·
Hello @Romulus !
Could you please provide us the steps to reproduce this error?
-
·
LeonidS
-
- · Romulus
- ·
The message comes from the sys_ai_log. I accessed the chat directly from the top header icon in the studio using GPT-4. I asked a few questions to test its responses. However, the chat wasn’t very smooth—it seemed to freeze at times, and the replies came in bursts. Because of this, I disabled it. I also noticed that the conversations were posted to the public outline/timeline. I was the only user on the site, which was running on a deicated server with an i5 processor, 8 cores, SSD, 64GB of RAM, and a 1Gb/s connection. The site itself loaded quickly and worked flawlessly, with memcached, nginx-proxy, and a Caddy reverse proxy in front. Despite this, the chat with ChatGPT was slow. Unfortunately, I can't reproduce the issue because I deleted the site; it was just a test upgrade from B2 to RC1, and the problem happened after the upgrade. Now I plan to do a clean install of RC1 to check its performance. It’s possible the issues were caused by the update, as there were other problems as well, so I can’t directly blame ChatGPT or the AI app. My custom config here:
custom Dockerfile:
FROM php:8.1-fpm RUN apt-get update && apt-get install -y \ cron \ libfreetype6-dev \ libjpeg62-turbo-dev \ libmcrypt-dev \ libpng-dev \ libonig-dev \ libmagickwand-dev \ libzip-dev \ unzip \ libmemcached-dev \ zlib1g-dev \ libssl-dev \ && docker-php-ext-configure gd --with-freetype --with-jpeg \ && docker-php-ext-install -j$(nproc) gd \ && docker-php-ext-install -j$(nproc) pdo pdo_mysql mysqli \ && docker-php-ext-install -j$(nproc) zip exif opcache iconv mbstring \ && pecl install xdebug && docker-php-ext-enable xdebug \ && pecl install mcrypt-1.0.3 && docker-php-ext-enable mcrypt \ && pecl install imagick-3.7.0 \ && pecl install memcached \ && docker-php-ext-enable imagick memcached
memcached.conf
upstream memcached_server { server 127.0.0.1:11211; # Alte servere Memcached pot fi adăugate aici, dacă este cazul }
rate_limit.conf
# rate_limit.conf limit_req_zone $binary_remote_addr zone=mylimit:10m rate=10r/s;
nginx default.conf
server { listen 9992 default_server; server_name localhost; root /opt/una; access_log /var/log/nginx/access.log; error_log /var/log/nginx/error.log; client_max_body_size 1024M; ########################################################################### ## Client Timeout ## ########################################################################### ## request timed out -- default 60 # read timeout for the request body from client, its set for testing purpose client_body_timeout 3600; # how long to wait for the client to send a request header, its set for testing purpose client_header_timeout 3600; # server will close connection after this time keepalive_timeout 3600; ## if client stop responding, free up memory -- default 60 send_timeout 3600; ## Reset lingering timed out connections. Deflect DDoS. ## allow the server to close connection on non responding client, this will free up memory reset_timedout_connection on; ########################################################################### ## Proxy Timeout ## ########################################################################### proxy_connect_timeout 3600; proxy_send_timeout 3600; proxy_read_timeout 3600; ########################################################################### ## Location / ## ########################################################################### location / { index index.html index.htm index.php; rewrite "^/page/(.*)$" /page.php?i=$1 last; rewrite "^/m/(.*)$" /modules/index.php?r=$1 last; rewrite "^/s/([a-zA-Z0-9_]+)/([a-zA-Z0-9\.]+)" /storage.php?o=$1&f=$2 last; if (!-e $request_filename) { rewrite ^/(.+)$ /r.php?_q=$1 last; break; } # Block requests with suspicious patterns
#if ($query_string ~* "(\%60|\%7C|\%26|\%24|\%3B|\%28|\%29)") {#return 403; #} } index index.php index.html index.htm; location ~ \.php$ { fastcgi_pass php:9000; fastcgi_param SCRIPT_FILENAME $document_root$fastcgi_script_name; include fastcgi_params; # Adăugați aceste setări de timeout fastcgi_connect_timeout 3600; fastcgi_send_timeout 3600; fastcgi_read_timeout 3600; fastcgi_buffer_size 128k; fastcgi_buffers 4 256k; fastcgi_busy_buffers_size 256k; fastcgi_temp_file_write_size 256k; } ########################################################################### ## HSTS policy ## ############################################################################add_header Strict-Transport-Security "max-age=31536000; includeSubDomains" always;#add_header X-Frame-Options "SAMEORIGIN" always;#add_header X-XSS-Protection "1; mode=block" always;#add_header X-Content-Type-Options "nosniff" always; # add_header Content-Security-Policy "default-src 'self'" always;#add_header Referrer-Policy "strict-origin-when-cross-origin" always; ########################################################################### ## serve static files directly Tanks to Alex T https://una.io/u/alex-t ## ########################################################################### location ~* ^(/cache_public/|/plugins_public/|/modules/|/studio/|/template/).+\.(jpg|jpeg|gif|css|png|js|ico|svg|eot|ttf|woff|woff2|)$ { access_log off; expires 1h; # root /opt/una # add_header Cache-Control "public"; try_files $uri =404; } ########################################################################### ## deny access to hidden files ## ########################################################################### location ~ /(\.ht|\.git) { deny all; } ########################################################################### ## deny access to specific folders ## ########################################################################### location ~ ^/(cache/|storage/|logs/|plugins/|tmp/) { deny all; } ########################################################################### ## Memcached Integration ## ########################################################################### location /memcached { # Example using just the URI as the key # set $memcached_key $uri; # Example with query parameters included # set $memcached_key $uri$is_args$args; # Example with a custom key set $memcached_key "key-$uri-$arg_user_id"; # Pass the key to the Memcached server memcached_pass memcached_server; # Define the response type default_type application/json; # Handle errors error_page 404 = /memcached_not_found; } location = /memcached_not_found { return 404 "{\"error\": \"Not found in Memcached\"}"; } ########################################################################### ## Permissions-Policy and CORS ## ############################################################################add_header Permissions-Policy "accelerometer=(), autoplay=(), camera=(), encrypted-media=(), geolocation=(), gyroscope=(), magnetometer=(), microphone=(), midi=(), payment=(), picture-in-picture=(), sync-xhr=(), usb=(), ch-ua-form-factor=('self' 'https://accounts.google.com')" always;#add_header Access-Control-Allow-Origin "https://www.googleapis.com https://accounts.google.com https://maps.googleapis.com https://www.coozila.com https://coozila.com https://agi.coozila.com https://onesignal.com" always;#add_header Access-Control-Allow-Credentials "true" always;#add_header Access-Control-Allow-Headers "Content-Type, Authorization, X-Requested-With" always; }my custom php.ini
; Memcached extension configuration [memcached] ; Memcached extension configuration ; Specify the serializer for Memcached memcached.serializer = php ; Specify the session prefix for Memcached memcached.sess_prefix = memc.sess.key. ; Enable binary protocol for Memcached sessions memcached.sess_binary = On ; Use SASL authentication for Memcached (0 = disable, 1 = enable) memcached.use_sasl = 0 ; Minimum waiting time for session lock acquisition (in milliseconds) memcached.sess_lock_wait_min = 150 ; Maximum waiting time for session lock acquisition (in milliseconds) memcached.sess_lock_wait_max = 150 ; Maximum size of POST data that PHP will accept. post_max_size = 1024M ; Maximum allowed size for uploaded files. upload_max_filesize = 1024M ; Maximum execution time of each script, in seconds max_execution_time = 3600 ; Maximum amount of time each script may spend parsing request data max_input_time = 3600 ; Maximum amount of memory a script may consume memory_limit = 1024M ; Maximum number of files that can be uploaded via a single request max_file_uploads = 50 ; Enable HTML error handling html_errors = On ; String to output before an error message error_prepend_string = " " ; String to output after an error message error_append_string = " " [opcache] ; Enables the opcode cache for PHP opcache.enable = 1 ; Frequency of checking for updated files (0 = never) opcache.revalidate_freq = 0 ; Validates cached files with timestamp checks opcache.validate_timestamps = 1 ; Maximum number of files that can be cached in memory opcache.max_accelerated_files = 100000 ; Memory consumption by the opcode cache (in MB) opcache.memory_consumption = 72 ; Maximum wasted memory percentage before restart opcache.max_wasted_percentage = 20 ; Size of the interned strings buffer (in MB) opcache.interned_strings_buffer = 16 ; Allows for faster shutting down of the opcode cache opcache.fast_shutdown = 1 display_errors = On error_reporting = E_ALL