Recent Questions - Server Fault |
- Windows 2019 and IIS Virtual SMTP Virtual Relay (Office 365)
- neg. lookahead not working for postfix header
- How can I migrate soft raid array made by mdadm to new server and new OS?
- Why windows Server 2008 R2 send/forward strong DNS requests or data?
- How to debug the cause of my suddenly extremely slow running database backups maintenance plan?
- Accessing dict subelement value
- Blocking a location in apache reverse proxy with require directives not working correctly
- How to host encrypted users' files in a way that I won't be able to read them?
- Obtaining the CPU L2 shared memory percentual of usage on Linux machine
- How should I set my nginx configuration to use subdomains with CNAME host record?
- Can I learn to RHCSA on EC2
- Configuring IPv6 to expose local device(s) to the internet
- Nginx Reverse proxy excluding files
- Why and how is Postfix automatically reading main.cf file?
- rdiff-backup on Windows: source from volume root, ignore $recycle.bin and "System Volume Information"
- How do I disable Bitlocker Encryption settings using Intune?
- OpenVPN fails to reconnect with wake-on-LAN after long sleep
- Windows 10 Group Policy Preferences Drive Maps failing over Wireless
- Credential Manager has been disabled by Administrator and cannot re-enable
- What's the easiest way to integrate Authy (two-factor) authentication with Apache httpd?
- Cisco IOS, Multiple WAN & Port Forwards (Outside -> Inside PAT)
- Random machine hangs with NFSv4 on CentOS/RHEL 6.5
- ssh tunnel refusing connections with "channel 2: open failed"
- Assign static IP to Users that access using Terminal Services (Remote Desktop)
- Nginx config reload without downtime
- Empty rewrite.log on Windows, RewriteLogLevel is in httpd.conf
- Cancel/Kill SQL-Server BACKUP in SUPSPENDED state (WRITELOG)
- Watchguard config, drop-in or mixed-routing mode?
- How to set guest user password on SAMBA server?
| Windows 2019 and IIS Virtual SMTP Virtual Relay (Office 365) Posted: 06 Sep 2021 09:04 PM PDT I have hit a wall with this setup and I cant for the life of me figure this out, despite having setup a few other Virtual SMTP servers in the past. Maybe something has changed in 2019? I will provide as much info as possible to help you help me :) I have installed SMTP Server and Telnet Client on my 2019 server, I have followed guides and rechecked my settings to ensure all is setup correctly. [General]
[Access]
[Messages]
[Delivery]
[LDAP] and [Security] tabs are default. The Office 365 user is set to SMTP auth enabled and I have confirmed this via powershell. telnet 127.0.0.1 25, shows: 220 mydomain.com Microsoft ESMTP MAIL Service, Version: 10.0.17763.1697 ready at Tue, 7 Sep 2021 13:53:35 +1000 However, when trying to send an email via this relay it stays in the C:\inetpub\mailroot\Queue directory and I get the below in the event log: Message delivery to the host 'X.X.X.X' failed while delivering to the remote domain 'recipientsdomain.com' for the following reason: The remote SMTP service rejected AUTH negotiation. I will probably kick myself once I figure this out, but I must be going mad. edit forgot to add... I can also telnet from the network to smtp.office365.com on port 587 and have also configured a connector in Office 365 for WAN IPs which are in use on my network. Thanks! Bil  |
| neg. lookahead not working for postfix header Posted: 06 Sep 2021 08:50 PM PDT I'm trying to exclude all emails not received from a mainstream domain The full postfix regex in /etc/postfix/header_checks is: Bu the negative lookahead is failing to match longer header strings. These longer header strings are seen in emails that are automatically forwarded from a gmail account (in the example below, forwarded from rocketman600@gmail.com to hello@mydomain.com): is it something to do with the  |
| How can I migrate soft raid array made by mdadm to new server and new OS? Posted: 06 Sep 2021 08:43 PM PDT I have a raid array which level is raid0 and made by mdadm, in my old server. The member of this array is 6 NVMe SSD, and I create this array with below coommand: Last week my old server OS disk (not this array) was broken. Since old server's IPMI have some dardware problem, I install a new OS in a new server, and plugin my all 6 NVMe SSD on new server. After reboot server, I found device /dev/dm0 is not exist. Then I tried execute below commadn: I wonder how can I find back my soft raid disk?  |
| Why windows Server 2008 R2 send/forward strong DNS requests or data? Posted: 06 Sep 2021 09:09 PM PDT My Server man has covid-19 and can not work because of this sorry if this is a simple question. I was checking my Server (Windows Server 2008 R2) and found that there is a high sending bit rate but no receiving (is zero always) to the URLs/Websites like below: It seems it is related to the DNS Rule(installed on my server) and when I stop this service, all data sending to the mentioned in the screenshot will stop. Is this normal or not and should I stop/block them? and how?
 |
| How to debug the cause of my suddenly extremely slow running database backups maintenance plan? Posted: 06 Sep 2021 07:20 PM PDT (Originally posted on DBA.StackExchange.com but closed, hopefully more relevant here.) Alexander and the Terrible, Horrible, No Good, Very Bad...backups. The Setup:I have an on-premise SQL Server 2016 Standard Edition instance running on a virtual machine from VMWare. @@Version:
The server itself is currently allocated 8 virtual processors, has 32 GB of memory, and all the disks are NVMes which get around 1 GB/sec of I/O. The databases themselves live on the G: drive, and the backups are separately stored on the P: drive. The total size across all of the databases is about 500 GB (before being compressed into the backup files themselves). The maintenance plan runs once a night (around 10:30 PM) to do a full backup on every database on the server. Nothing else out of the ordinary is running on the server, nor is anything else running at that time in particular. The Power Plan off the server is set to "Balanced" (and "Turn off hard disk after" is set to 0 minutes aka never turn it off). What Happened:For the last year or so, the total runtime for the maintenance plan job took about 15 minutes total to complete. Since last week, it has skyrocketed to taking about 40x as long, about 15 hours to complete. The only thing I'm aware of changing on the same day the maintenance plan slowed down was the following Windows updates were installed on the machine prior to the maintenance plan running:
We also have another similarly provisioned SQL Server instance on another VM that underwent the same Windows updates, and then subsequently experienced slower backups as well after. Thinking the Windows updates were directly the cause, we rolled them back completely, and the backups maintenance plan still runs extremely slow anyway. Weirdly, restoring the backups for a given database happens very quickly, and uses almost the full 1 GB/sec of I/O on the NVMes. Things I've Tried:When using Adam Mechanic's sp_whoisactive, I've identified that the Last Wait Types of the backup processes are always indicative of a disk performance issue. I always see
When looking at the Resource Monitor on the server itself, during the backups, the Disk I/O section shows that the total I/O being utilized is only about 14 MB/sec (the most I've ever seen since this issue occurred is 30 MB/sec):
After stumbling on this helpful Brent Ozar article on using DiskSpd, I tried running it myself under similar parameters (only lowering the number of threads to 8 since I have 8 virtual processors on the server and setting the writes to 50%). This is the exact command
The DiskSpd results seem literally unbelievable. After further reading, I stumbled on a query from Paul Randall that returns disk latency metrics per database. These were the results:
The worst Write Latency was 63 milliseconds and the worst Read Latency was 6 milliseconds, so that seems to be a big variance from DiskSpd, and doesn't seem terrible enough to be the root cause of my issue. Cross-checking things further, I ran a few PerfMon counters on the server itself, per this Microsoft article, and these were the results:
Nothing extraordinary here, the max value of all the counters I measured was 0.007 (which I believe is milliseconds?). Finally, I had my Infrastructure team check the disk latency metrics that VMWare was logging during the backups job and these were the results:
Seems like at worst, there was a spike of latency of about 200 milliseconds around midnight, and the highest I/O was 600 KB/sec (which I don't really understand since the Resource Monitor is showing that the backups are at least using around 14 MB/sec of I/O). Other Things I've Tried:I just tried restoring one of the larger databases (it's about 250 GB) and it only took about 8 minutes total to restore. Then I tried running
Here is sp_whoisactive results when I first ran Start: 5% Done: I'm guessing this is normal with it just being an estimate, and 16 minutes doesn't seem too bad for a 250 GB database (though I'm not sure if that's normal) but again the I/O was only maxing out at about 10% of the drive's capabilities, with nothing else running on the server or the SQL instance. These are the results of I also have been experiencing weird slowness issues with the
About 5 minutes later, and it's still stuck at the same percent complete, and my Transaction Log Backups (that usually finish in 1-2 seconds) have been under contention for about 30 seconds:
15 minutes later and the
Then I got an error with the
I retried Then I tried manually scripting a T-SQL backup to a file on the P: drive and that ran slow just like the maintenance plan backup job:
I ended up cancelling it after about 3 minutes, and it immediately rolled back. Summary:Coincidentally, the backups maintenance plan job got about 40x slower (from 15 minutes to 15 hours) every night, right after Windows updates were installed. Rolling back those Windows updates didn't fix the issue. SQL Server Wait Types, Resource Monitor, and Microsoft DiskSpd indicate a disk problem (I/O in particular), but all other measurements from Paul Randall's query, PerfMon, and VMWare Logs don't report any issues with the disks. Restoring the backups for a particular database are quick and use almost the full 1 GB/sec I/O. I'm scratching my head...  |
| Accessing dict subelement value Posted: 06 Sep 2021 07:17 PM PDT Here's my playbook And the output for this is I need to get the luns of a matched vg I have tried this beetwen other ways VG is an extravariable where I put the matched VGname Hope you can help  |
| Blocking a location in apache reverse proxy with require directives not working correctly Posted: 06 Sep 2021 06:03 PM PDT so I have a virtual host reverse proxy pointing to a service at servicesubdoamin.mydomain.com and that service has an admin page at servicesubdoamin.mydomain.com/admin and I want to block access to the admin page and only allow it from local ip's. I've tried a couple different things that should work however when I try to access them from outside my local netowrk the page gets served with no issue. This is what I currenty have: but I have also tried and also placing the But nothing works. I cant seem to block this page, any and all help is greatly appreciated.  |
| How to host encrypted users' files in a way that I won't be able to read them? Posted: 06 Sep 2021 08:40 PM PDT I am making plans for a new project and looking for a way to allow users upload their files and keep them encrypted and only accessible by them, but at the same time I need their contacts to access the files too. To explain a little bit more:
Is something like that even possible?  |
| Obtaining the CPU L2 shared memory percentual of usage on Linux machine Posted: 06 Sep 2021 04:52 PM PDT on linux based systems, how can I estimate or maybe read the CPU L2 shared memory % of usage?  |
| How should I set my nginx configuration to use subdomains with CNAME host record? Posted: 06 Sep 2021 06:06 PM PDT running Debian 11. How should I add my subdomain to my site using my CNAME DNS records? My website directory is located at /var/www/ and I'm wondering if I can make a subdirectory for the subdomain within that directory. i.e. /var/www/domain-dir/subdomain-dir Can I add subdomain server_name to right in my sites-available config file? I don't want to create tons of A/AAAA records to point to my subdomain as it could become messy as I'll likely create more subdomains later, blog.example.com, git.example.com, etc.. Sorry if this post is long, I'm new to this and don't know how to make it more concise!  |
| Posted: 06 Sep 2021 05:41 PM PDT I have MacBook with M1 processor and I want to learn to RHCSA exam. There are no ARM distributions yet and I can't create local environment, would it be ok to provision EC2 centOS instance? I'm just afraid that I won't be able to recreate some test cases.  |
| Configuring IPv6 to expose local device(s) to the internet Posted: 06 Sep 2021 08:58 PM PDT I am trying to expose a local client to the net to host a website. I am struggling to understand IPv6. Current setup: I've configured the TP-Link router to use IPv6. In the router's menu I see: The "global address" under "IPv6/WAN" is The "LAN IPv6 address" under "IPv6/LAN" is My questions:
 |
| Nginx Reverse proxy excluding files Posted: 06 Sep 2021 09:02 PM PDT I've a landing page done using wordpress and is hosted on example.com. I've an app running on app.example.com on external url. When user try to access wordpress files, it should be served from example.com and if that url or folder is not available the url must be masked and must go to remote url as example.com/$1. I've tried using nginx reverse-proxy but it is not working.  |
| Why and how is Postfix automatically reading main.cf file? Posted: 06 Sep 2021 05:19 PM PDT today I changed TLS cert paths in Postfix main.cf file. Those paths turned out to be broken. In a few minutes, Postfix has read those changed paths and my TLS connections became broken (becouse of broken paths). Question 1: Why Postfix automatically reads changes from main.cf file? Is is documented? I can't find any information about it. Question 2: Can I turn off this behaviour? I expected Postfix to replace certs after reload of the service, not on the fly. Thanks  |
| Posted: 06 Sep 2021 05:10 PM PDT I'm trying to use rdiff-backup in Windows, using the root of a local disk as a source and a local folder on a different local disk as a destination. rdiff-backup keeps crashing when it tries reading the "System Volume Information" folder; I am trying to discover what syntax may cause it to ignore that one, and "$RECYCLE.BIN" as well; I tried giving an ignore file with those absolute paths in it, with paths with those prefixed with ** to try and match any files like that, ran an experiment to see if it would ignore those if I created them in a test folder (which succeeded), etc. So, is there a syntax that will cause rdiff-backup to ignore those in the root of a mounted drive? As a starting point I tried: but had no joy.  |
| How do I disable Bitlocker Encryption settings using Intune? Posted: 06 Sep 2021 05:34 PM PDT We've activated Intune Bitlocker encryption and configured it needs a password to unlock. Since we don't want our users to change the Bitlocker pin, we want to disable the Settings below. For all non Germans, it's under: Thanks for any help! :)  |
| OpenVPN fails to reconnect with wake-on-LAN after long sleep Posted: 06 Sep 2021 09:06 PM PDT I have a RaspberryPi on my parents' house with PiVPN set up and configured to provide a personal VPN service for me and a few friends. This VPN has worked flawlessly since the beggining, I have used it with my PC and never got an error. I recently set up another computer with Windows10 at my parents' house, to act as a server for various purposes (in case it is related to this issue, I use it as a home multimedia server with Plex Media Server and also as a Git Repository for personal use). I need it to connect automatically to the VPN, so I did the following:
Now the problem appeared when I tested the VPN connection to the server after the "2AM sleep". I woke up the server and then tried to ping it as usual with its static VPN IP but I couldn't reach it. I logged in through TeamViewer to check what was happening and when I opened OpenVPN's gui, I found that it was stuck in a loop like this: I tested the VPN with my PC and works nicely as usual, so the best bet is that it's the server's fault. I personally think that maybe has something to do with the batch script I made and programmed to run at 2AM to put the PC to sleep at 2AM, because I had no problems with other sleep methods (manual sleep and inactivity sleep). The batch script looks like this: I used this script because I saw a tutorial on how to do a batch script for this. As in that tutorial said, I also ran the following command in order to do sleep instead of hibernation: What could be the problem?  |
| Windows 10 Group Policy Preferences Drive Maps failing over Wireless Posted: 06 Sep 2021 08:08 PM PDT I'm struggling getting Group Policy Preferences Drive Mapping to work over wireless (WPA2-Enterprise using Certificates) from our Windows 10 Surface Pro 4s. The Active Directory user account's Home Folder drive map also does not appear. All of these paths use DFS (Server 2008 R2). Shortly after login, a manual Gpupdate will cause the mapped drives to appear. Waiting 30 seconds before login also works. We've had the "Always wait for the network at computer startup and logon" enabled since XP days. I tried setting the "Specify startup policy processing wait time" to 60 but this made no difference (nor did it lengthen boot). The wireless NIC does not appear to have a "Wait For Link" type setting to enable. Event logs show Event ID 4098 with source "Group Policy Drive Maps" saying the preference item "failed with error code '0x80070035 The network path was not found.'" I had wondered if the underlying problem might be the new UNC Hardening feature but even adding an exception for "\\DomainNetBIOSname" did not help. (See here: Windows 10: Group Policy fails to apply directly after boot, succeeds later) The only significant clue to what's going on is that when I changed my user account home folder to a direct UNC path to the server rather than via DFS, my home drive was able to appear correctly. The DFS Client service (as seen in regedit) already has a Start type signifying "System". I'm not sure where to go from here. Does anyone have any ideas? Thanks!  |
| Credential Manager has been disabled by Administrator and cannot re-enable Posted: 06 Sep 2021 06:03 PM PDT I'm having trouble getting a Windows 7 work computer to keep credentials for a mapped drive. Whenever I load Credential Manager, I get "Windows Credentials have been disabled by your Administrator." I'm logged into an Administrator account (but not the built-in account), and haven't had much luck removing this restriction. The computer is not on a domain, so there is no domain policy enforcing this. I checked gpedit.msc, and looked under:
I found the option Network Access: Do not allow storage of passwords and credentials for network authentication, but it is already disabled. I performed this check on both the server and user computer. I made sure it is in the same work group as all of the other computers. I tried CCleaner to see if maybe there was a registry issue, but it just found missing .DLLs and unnecessary file types. I was trying VaultCMD in Command Prompt, and created a new Vault, but I don't have the proper option to store the credentials for the server. Any suggestions? Thanks in advance.  |
| What's the easiest way to integrate Authy (two-factor) authentication with Apache httpd? Posted: 06 Sep 2021 07:06 PM PDT EDIT: Something similar to Authy would work too, if that service was i) hosted/SaaS and ii) able to send SMS messages.
(*Apache httpd may be replaced by NGINX at some point, so ideally the solution suggested would carry over, but please don't refrain from suggesting Apache httpd-only solutions!)  |
| Cisco IOS, Multiple WAN & Port Forwards (Outside -> Inside PAT) Posted: 06 Sep 2021 04:04 PM PDT I have been trying to work out how to accomplish PATing from Outside to Inside on a Cisco IOS router, in this case specifically a Firstly, the problem I am trying to solve is that we'd like external port forwards to work regardless of which connection is the default gateway. An example port forward rule in this device: Where This works fine if It's also worth noting that the I am performing my standard Inside to Outside PAT by using I know I can mess around with The only reason why I think this is my best bet is the fact that every firewall device I've used has functionality in its policies to perform source NAT translation to the IP address of the egress interface, and it is so simple to turn on! Edit: Watered down config |
| Random machine hangs with NFSv4 on CentOS/RHEL 6.5 Posted: 06 Sep 2021 08:08 PM PDT We have an in-house "compute farm" with about 100 CentOS (free re-distribution of RHEL) 5.7 and 6.5 x86_64 servers. (We are in the process of upgrading all the 5.7 boxes to 6.5.) All these machines do two NFSv4 mounts (with sec=krb5p) to two CentOS 6.5 servers. One NFS server is for user home directories, the other contains various data for user processes. Randomly, one of the client machines will get into a bad state such that any access to the NFSv4 mount hangs ("ls" for example). This means no one (except root) can login, and all user processes that require access to the shares get stuck. In other words, so far this is non-deterministic and cannot be replicated. I have very verbose NFS logging enabled in both the clients and servers, but never get any errors. However, when this state is triggered, I do get these kernel trace errors on the client machines: At this point, the only reliable way to make the machine usable again is to reboot it. (And even that requires a hard power cycle, since the software reboot hangs when it tries to unmount the NFS filesystems.) It seems like this problem is correlated with a process that malfunctions and starts writing data like crazy. For example, a segfault that generates a huge core file, or a bug with a tight print loop. However, I've tried to duplicate this problem in a lab environment with multiple "dd" processes hammering away at the NFS server, but all machines chug along happily.  |
| ssh tunnel refusing connections with "channel 2: open failed" Posted: 06 Sep 2021 07:31 PM PDT All of a sudden (read: without changing any parameters) my netbsd virtualmachine started acting oddly. The symptoms concern ssh tunneling. From my laptop I launch: Then, in another shell: The ssh debug says: I tried also with localhost:80 to connect to the (remote) web server, with identical results. The remote host runs NetBSD: I am a bit lost. I tried running I tried restarting the ssh daemon to no avail. I haven't rebooted yet - perhaps somebody here can suggest other diagnostics. I think it might either be the virtual network card driver, or somebody rooted our ssh. Ideas..?  |
| Assign static IP to Users that access using Terminal Services (Remote Desktop) Posted: 06 Sep 2021 05:00 PM PDT For a project I'm working on I'm looking for the ability to assign a specific ip to users when they start a terminal service session. I'm using Windows Server 2008 R2 SP1 and I tried using Remote Desktop IP Virtualization but as far I understand it lets me only to enable the ability to assign a random ip to a session when the user access with or without a dhcp (changing some registry keys). I need this to set up filtering rules per user on the project firewall. EDIT From what I understand there are a couple of dlls that handle (in "fake dhcp" mode, changing the registry keys) the ip assigment. If assign a static IP to an Users isn't actually supported, can a library be built from scratch to handle this situation and, if yes, when i can find some MS docs about these libraries (I refer to TSVIPool.dll and the second one that can be assigned to the key Control in the same registry path, I can't find the name)  |
| Nginx config reload without downtime Posted: 06 Sep 2021 06:20 PM PDT I use nginx as a reverse proxy. Whenever I update the config for it using I face a brief downtime. How can I avoid that?  |
| Empty rewrite.log on Windows, RewriteLogLevel is in httpd.conf Posted: 06 Sep 2021 06:03 PM PDT I am using mod_rewrite on Apache 2.2, Windows 7, and it is working ... except I don't see any logging information. I added these lines to the end of my httpd.conf: The log file is created when Apache starts (so it's not a permission problem), but it remains empty. I thought there might be a conflicting What else could cause this? Could this be caused by Apache not flushing the log file? (I closed it by hitting CTRL-C on the httpd.exe command ... this caused the access logs to be flushed to disk, but still nothing in rewrite.log) My (partial) httpd-vhosts.conf:  |
| Cancel/Kill SQL-Server BACKUP in SUPSPENDED state (WRITELOG) Posted: 06 Sep 2021 04:04 PM PDT I have a SQL 2008 R2 Express on which backups are made by executing sqlmaint from windows task planer. Several backups ran into an error and got stuck in state How can I get these backup processes to stop so they release resources? Simply killing the processes doesn't work. The process will stay in  |
| Watchguard config, drop-in or mixed-routing mode? Posted: 06 Sep 2021 05:00 PM PDT I have a Watchguard XTM 2 that is currently acting as a firewall and a router for my business network, I currently have the WG setup in mixed-routing mode and am happy with the current configuration. The reason I am curious about drop-in mode is because I would like to use all the interfaces on the back of the watchguard for the same subnet. My understanding is that drop-in mode will put them all on the same subnet, but it is unclear from the manual that the routing/firewall/vpn will still work as expected. This WG is right behind a DSL modem that is setup in bridge mode, so the WG is handling all PPPoE auth and routing for the network.  |
| How to set guest user password on SAMBA server? Posted: 06 Sep 2021 07:06 PM PDT I'd be generally ok with guest and an empty password as I don't need any access rights management among my users. But the server is a remote internet-accessed machine, so I'd prefer to set a good password for it. What's the most simple way? Can I just specify a password in samba.conf, or absolutely need to use LDAP or add users to the server system?  |
| You are subscribed to email updates from Recent Questions - Server Fault. To stop receiving these emails, you may unsubscribe now. | Email delivery powered by Google |
| Google, 1600 Amphitheatre Parkway, Mountain View, CA 94043, United States | |
No comments:
Post a Comment