serverfault.com

I had a few servers in the Bahrain region and now that it is completely unavailable, I don't know what to do next. The snapshots of the volumes o. had as well are in the same region. The web GUI of AWS console returns timeout error, I tried accessing the snapshots via CLI which as well fails due to region not being accessible. Are there any other options to get hand son the data from this region?

I'm fighting an issue with my 4-way Ceph cluster on Ubuntu 24.04.4 LTS: Two nodes (blade3n1 and blade3n2) already work fine, but the other two (blade3n3 and blade3n4) do not accept daemon placements despite being listed as hosts: mixtile@blade3n1:~$ sudo ceph orch ps NAME HOST PORTS STATUS REFRESHED AGE MEM USE MEM LIM VERSION IMAGE ID CONTAINER ID alertmanager.blade3n2 blade3n2 *:9093,9094 running (2h) 2m ago 3d 23.9M - 0.25.0 fd4d9a3dd1fd c7d1f967aa50 ceph-exporter.blade3n1 blade3n1 running (2h) 2m ago 3d 11.7M - 19.2.3 af0c5903e901 5ad19356bf5f ceph-exporter.blade3n2 blade3n2 running (2h) 2m ago 3d 6716k - 19.2.3 af0c5903e901 4e020baf8e31 crash.blade3n1 blade3n1 r

My home server is running postfix, and is configured to send out monitoring emails (I don't need to receive there). This worked fine until I installed OpenVPN. Whenever the VPN is ON, the server can reach anything on the Internet, but emails are queued. If I turn off the VPN and flush the queue, everything is sent and I can see the alerts in my mailbox. I reached out to my VPN provider - ProtonVPN - and they confirmed they are not doing anything to block email traffic. So there is something wrong with my postfix configuration and I need help pinning it down. I see these entries in my logs when the VPN is ON root@home-server:~# journalctl -r -t postfix/smtp Apr 19 11:17:00 home-server postfix/smtp[482862]: 2FDEA113: to=<[email protected]>, orig_to=<root>, relay=mail.remote-server.com[62.XXX.XXX.XXX]:25, delay=252420, delays=252400/0.03/20/0, dsn=4.4.2, status=deferred (lost connection with mail.remote-server.com[62.XXX.XXX.XXX] whi

We have at the moment an Openshift 4.21 cluster where all networks are in the default namespace and all the VMS as well. The "Migration Toolkit for Virtualization" is in the namespace openshift-mtv. We setup the network with this guide. Plus as an test we created an "Virtual Machines Networks". And we setup and configure the migrations from vmware with this guide. When we want to create the network map, we have the problem that we can not see the networks. How can we configure that the toolkit can see the networks?

Here yet another production webserver trying to survive AI-slop-aggressive-web-crawlers-botnets, by banning IPs. Unuseful Preamble Most of the time, the production webserver has very decent CPU load, but in very specific milliseconds I'm hit by ~150 visits by ~150 different IPs on the very specific page (and minor variants) and all resources, in that millisecond, are saturated. As usual, these aggressive web crawlers: do not respect robots.txt do not respect rel="nofollow" they just act like DDoS In this server I have Apache HTTPd and I've no intention for many reasons to put Anubis in front of my webserver. Mostly because I believe in the web, I don't want to enshittificate the web, and also I do not want to introduce serious accessibility regressions for my visitors who want to use

I have an IPv6 server and client on a local network. How do I test it outside of the local network? I tried to test it from an IPv4 network, but it doesn't work because I am in front of a CGNAT. I want to test if it could receive packets from other computers outside my network. import socket # Use AF_INET6 for IPv6 and SOCK_STREAM for TCP server = socket.socket(socket.AF_INET6, socket.SOCK_STREAM) # Bind to all interfaces (::) on port 8080 # IPv6 address tuples can have up to 4 elements: (host, port, flowinfo, scope_id) server.bind(('::', 25010)) server.listen(5) print("Server listening on port 25010..") while True: # Accept a new connection conn, addr = server.accept() print(f"Connected by {addr}") while True: # Receive up to 1024 bytes data = conn.recv(1024) if not data: break # Print the decoded message print(f"Rece

I have a pretty basic postfix + dovecot setup that's working to deliver email from postfix to dovecot using LDA. I'm trying to switch to LMTP because the documentation says it's supposed to be better for various reasons, but I keep getting hit with the 5.1.1 "User doesn't exist" error upon delivery whatever I try: Apr 08 11:55:05 vps postfix/smtps/smtpd[26461]: connect from 82-64-153-141.subs.proxad.net[<IP_REDACTED>] Apr 08 11:55:05 vps postfix/smtps/smtpd[26461]: 848E24F4E8: client=82-64-153-141.subs.proxad.net[<IP_REDACTED>], sasl_method=PLAIN, sasl_username=user Apr 08 11:55:05 vps postfix/cleanup[26466]: 848E24F4E8: message-id=<[email protected]> Apr 08 11:55:05 vps postfix/qmgr[26446]: 848E24F4E8: from=<[email protected]>, size=575, nrcpt=1 (queue active) Apr 08 11:55:05 vps postfix/smtps/smtpd[26461]: disconnect from 82-64-153-141.subs.proxad.net[<IP_REDACTED>] ehlo=1 auth=1 mail=1 rcpt=1 dat

I have a code-signed desktop application distributed to thousands of clients with embedded HTTP Public Key Pinning (HPKP) for latercheck upon a new release. Our DigiCert code-signing certificate expired, and we purchased a new certificate. However, the new certificate has a different public key, causing HPKP validation to fail and breaking our auto-update mechanism. The old certificate private key is on a SafeNet eToken 5100 (Gemalto ID Prime MD) that is now PIN-locked (user PIN locked, user PIN count low). A former colleague who initialized the token has left, and we lost the user PIN. What I need to know: Can I replace the HPKP of the new certificate with the HPKP of the old certificate? Does unlocking with SO-PIN preserve the old private key? Will the old private key remain

Our website uses an SSL certificate from Sectigo. It works fine except for browsers on old devices: https://www.ssllabs.com/ shows no chain issues: However, https://www.sslchecker.com/sslchecker shows missing CHAIN CERT 1, CHAIN CERT 2, and ROOT 1. Could anyone offer a tip on how to remedy this? Update I added the following

When installing SQL Server 2022 (16.x), is there a command line argument or configuration file option that can skip the Azure Extension for SQL Server step, or at least leave the Azure Extension for SQL Server checkbox unchecked? I thought /FEATURES=SQLEngine (since the extension is installed by /FEATURES=AZUREEXTENSION) would do the trick, but Azure Extension is still enabled by default.