serverfault.com

I'm using a local server (Windows server 2022 standard) where MariaDB 10.11.6 is currently running. My modem is directly connected to the fiber network with a Draytek Vigor 2135F. In this router I blocked all access to MariaDB access and granted only my websites. So far, so good.. But the sites which connect to my DB are really slow. In my mariaDb ini I have this: [mysqld] datadir=C:/Program Files/MariaDB 10.11/data port=3306 host_cache_size = 1000 innodb_buffer_pool_size = 2G innodb_read_io_threads = 8 innodb_thread_concurrency = 16 innodb_io_capacity = 2000 innodb_write_io_threads = 8 innodb_log_file_size = 256M innodb_log_buffer_size = 8M join_buffer_size = 2M sort_buffer_size = 2M key_buffer_size = 32M max_connections = 500 max_user_connections = 500 max_allowed_packet = 16M max_heap_table_size = 32M myisam_sort_buffer_size = 64M read_buffer_size = 8M read_rnd_buffer_size = 32M thread_cache_size = 100 tmp_table_size = 64M wait_timeout = 160 sort_buffer_size

I have a Gluster volume that was originally on server A, and then was configured to be replicated on server B. By mistake, I too apart server A before everything was replicated to server B. Now, server B shows some files as unavailable. Here, 2.tar.gz is shown in a form of question marks, as it wasn't replicated. root@serverB:/mnt/example# ll -h ls: cannot access '/mnt/example/2.tar.gz': No data available total 16G drwxrwxr-x 2 root root 4.0K Feb 21 22:53 ./ drwxr-xr-x 11 root root 4.0K Sep 17 2021 ../ -rw-r--r-- 1 root root 3.0G Sep 8 2011 1.tar.gz -????????? ? ? ? ? ? -r--r--r-- 1 root root 4.9G Oct 8 2015 3.tar.gz ... I don't care much about the actual missing files—I have a backup. I would like, however, to get rid of those non-existent files. An rm doesn't work, because: root@serverB:/mnt/exam

Running redmine under apache 2.4 on ubuntu 24.04. The apache server is set up to require logins. Under normal circumstances, one logs in to the apache server once, then when one requests a redmine page one logs into redmine once. Subsequently, additional tabs require no logins on either apache or redmine. However, under some conditions which I have not been able to pinpoint, when one switches browser tabs or tries to update an issue, the apache server asks for another login; redmine seems to maintain state ok and does not request additional logins. When this happens the login request seems like it is in some kind of loop, as it comes up multiple times. It's not clear whether it is asking for logins for tabs other than the one currently being viewed and is getting confused, or if it is just in a loop. In order to get out of this condition I have had to completely exit the browser. It happens with both firefox and chrome. Does anyone have any idea wh

Since several days now, my Linux desktop (Fedora Linux 43 Workstation) has occasional IO "hangs". So far I noticed Playing MP3s from an NFS filemount Writing to the local filesystem (e.g. when saving a file in a texteditor) Reading from the local filesystem When it happens, the thread with the operation is freezing 2-20 seconds and then continues normally. I don't see any messages journalctl or dmesg. I suspect that it started with the last batch of installed package updates, but I have no clue how to approach this without a hint in the logs. Current kernel is 6.18.8-200.fc43.x86_64, built on 30. January (not by me). Do you have any other advice how to triage this?

I bought a WiFi card on Amazon for an Ubuntu Server. I know it's not optimal for a server, but I don't need it running 24/7—it's for personal use and I can't run a cable. I installed the card, and when I run ip a, I can see the new WiFi interface: 3: wlp2s0: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc noqueue state DOWN group default qlen 1000 link/ether e0:e2:58:aa:13:51 brd ff:ff:ff:ff:ff:ff I modified my /etc/netplan/01-netcfg.yaml config file to manage this new interface. But when I run sudo netplan generate followed by sudo netplan apply, I still don't get an IP address when I check ip a. network: version: 2 renderer: networkd ethernets: enp3s0: dhcp4: true nameservers: addresses: [8.8.8.8, 1.1.1.1] wifis: wlp2s0: dhcp4: true access-points: "Freebox-172BD1": password: password nameservers: addresses: [8.8.8.8, 1.1.1.1] I suspect I need to install a driver, but the guide that came with the card only mentions Windows and insert

I need to change the timestamp format of the logs in C:\Windows\System32\dns\dns.log to include the complete year in the logs timestamp. Since the time stamp format in this log file is based on Region settings, I would have to change the format there and then use the "Administrative" tab in the "Region" settings to Copy settings to the system account. I believe this not only changes the format in dns.log but also systemwide. Since this is a production Domain controller, I would like to know what adverse effects this could have. Will it affect the current functionality of the domain controller? If this not recommended what other alternative method is possible to just change the format only in the dns.log?

I have multiple computers all connected through WiFi except a raspberry pi connected through ethernet that have always been able to communicate with each other over SSH, HTTP for many years. But suddenly, one of them was unable to connect to or receive connections from all the others except my raspberry pi connected over ethernet. In my router/modem (Valerie from Virgin (Bell)), I have three radios 2.5GHz, 5GHz radio 1, 5GHz radio 2. I turned off all but 5GHz radio 1 and that allowed everyone to start talking to each other. The machine that was "isolated" is a Mac Mini that I used to host a personal website and a few other web doohickeys. It had been off and unplugged since last year but I decided to turn it back on and setup my website and stuff. I wanted to work with it from my main machine over SSH but I couldn't SSH to it, and it couldn't SSH to my main machine either. I want to understand how I could have figured out what the prob

I recently upgraded a backup server using bacula to Ubuntu 24.04. After this upgrade the machine failed to reboot properly. Watching the boot sequence it was complaining the the L2/L3 cache of the RAID controller was not working and suggested replacing the RAID controller. I did so and now by removing one of the entries in /etc/fstab the server does reboot and allow login. However, it no longer able to access and configure the RAID controller using the storcli64 program. Here is the output: sudo /opt/MegaRAID/storcli/storcli64 show CLI Version = 007.3603.0000.0000 Oct 30, 2025 Operating system = Linux 6.8.0-100-generic Status Code = 0 Status = Success Description = None Number of Controllers = 0 Host Name = ohr Operating System = Linux 6.8.0-100-generic The controller hardware itself is visible on the pci bus: 07:00.0 RAID bus controller: Broadcom / LSI MegaRAID SAS-3 3008 [Fury] (

I’m trying to set up an OpenLDAP proxy so that: Identity + password come from an external corporate LDAP (ou=Users,o=abc) POSIX attributes (uidNumber, gidNumber, homeDirectory) are stored locally SSSD queries a single LDAP endpoint (our IAM LDAP) with merged data of posix and user Password is not exposed by client, so bind forwarding needed for that Architecture SSSD → IAM OpenLDAP (proxy-readonly) -> IAM OpenLDAP(local POSIX) → External LDAP (identity + auth) local rootDN : dc=iam,dc=abc,dc=com, External DN : ou=Users,o=abc (This is dynamic and client can provide any URI and DN) Our IAM LDAP runs inside Kubernetes using this image: jpgouin/openldap:2.6.9-fix ------------Approached I tried------------------- Attempt #1 — back‑meta Created olcDatabase=meta olcDbRewrite: {0}suffixmassage ou=users,dc=i

We get this error starting this morning, when accesing to Google App Engine Cloud Console, Something went wrong. There was an error while loading /appengine?project=plenion-cloud. Please try again. It may be a browser or network issue. Go to the loading issues help page to troubleshoot the issue. Request ID: 7831925681237543667 Also from Google Cloud CLI, we get 500 erros.