News items for tag homeserver - Koos van den Hout

2022-05-09 Grafana alerts working again
After reverting to Grafana 8.4.7 for a while because alerts were failing in Grafana 8.5.0 I had a look at the available version today and saw version 8.5.2. I assumed the problem with DataSourceNoData errors was fixed by now and did the upgrade.

Indeed the alerts are seeing data fine now and I trust they will work when needed.

Tags: , ,
2022-04-23 Grafana alerts failing in 8.5.0
I installed Grafana from their debian repository, so I get updates via the normal apt update / apt dist-upgrade process. Since upgrading to version 8.5.0 the alerts were all firing because of 'DatasourceNoData' errors. According to Alert Rule returned no data (after upgrade to 8.5.0) #48128 other people are seeing this too.

For now I downgraded to version 8.4.7 where things work fine and I'll see if a newer version shows up.

Tags: , ,
2021-12-27 Raid-1 on the homeserver rebuilt
After seeing read errors on one disk in the raid-1 of the homeserver I ordered a replacement SSD of a different brand and exactly the same size. It arrived today, and I did the work to replace the suspect disk.

First set the old disk as failed and removed from the array. And note the complete serial number on a piece of paper to make sure I removed the faulty disk.

After that the server was shut down, disconnected from a lot of cables, dragged from the homerack in the attic and I worked on it. It took a while to open the side with the SSDs (below the mainboard) and with two exactly the same SSDs it was a 50% chance which one to remove. After removing the disk tray and unscrewing the SSD from the disk tray I was able to read the physical label on the underside and I guessed right.

After that the new disk was installed, the case closed again and dragged back to its place and cables connected again. After boot it came all up fine.

After bootup I partitioned the new disk, added it to the raid-1 again and set up the EFI and Linux boot partitions on the disk.

Last step was to setup the boot menu with efibootmgr to set both disks as bootable.

Tags: , ,
2021-12-21 New ssd for the homeserver ordered
I noticed syslog messages I don't like:
[17200683.290921] md: data-check of RAID array md127
[17200683.291277] md: minimum _guaranteed_  speed: 1000 KB/sec/disk.
[17200683.291619] md: using maximum available idle IO bandwidth (but not more than 200000 KB/sec) for data-check.
[17200683.291935] md: using 128k window, over a total of 937253184k.
[17201245.784689] ata2.00: exception Emask 0x0 SAct 0x1fe00000 SErr 0x0 action 0x0
[17201245.785175] ata2.00: irq_stat 0x40000008
[17201245.785465] ata2.00: failed command: READ FPDMA QUEUED
[17201245.785766] ata2.00: cmd 60/80:a8:00:52:51/00:00:0c:00:00/40 tag 21 ncq dma 65536 in
                           res 41/40:20:60:52:51/00:00:0c:00:00/00 Emask 0x409 (media error) <F>
[17201245.786402] ata2.00: status: { DRDY ERR }
[17201245.786737] ata2.00: error: { UNC }
[17201245.787281] ata2.00: configured for UDMA/133
[17201245.787619] sd 1:0:0:0: [sdb] tag#21 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE
[17201245.787966] sd 1:0:0:0: [sdb] tag#21 Sense Key : Medium Error [current] 
[17201245.788317] sd 1:0:0:0: [sdb] tag#21 Add. Sense: Unrecovered read error - auto reallocate failed
[17201245.788689] sd 1:0:0:0: [sdb] tag#21 CDB: Read(10) 28 00 0c 51 52 00 00 00 80 00
[17201245.789123] blk_update_request: I/O error, dev sdb, sector 206656096
[17201245.789530] ata2: EH complete
And a number of other errors on sdb. Time to replace it! I ordered a new ssd. This time a different brand. Current configuration is with 2 Kingston drives with very close serial numbers, so maybe the other drive will give similar issues soon.

The check of the raid1 mirror was also showing differences. I'm waiting for the replacement ssd to show up, and at that moment I will remove the suspect ssd from the array and replace it.

Update 2021-12-24: Writing about the order helped speed things up: I just received notification the replacement ssd is being sent. Which will not show up until after Christmas. I also noticed the problematic Kingston still has warranty, so maybe I can get a replacement for that one too. They came in about 1.5 years ago when I upgraded the storage on the homeserver.

Tags: , ,
2021-10-22 Naming interfaces used by libvirt virtual machines
The homeserver conway has an ever growing list of network interfaces, also due to adding a DMZ network.

This was starting to look a bit messy, with things like:
koos@conway:~$ /sbin/brctl show brwireless
bridge name     bridge id               STP enabled     interfaces
brwireless              8000.4ccc6a8efa4b       no              enp10s0.3
Solution: name the interfaces in the VM definitions, like:
    <interface type='bridge'>
      <source bridge='brdmz'/>
      <target dev='dmz-minsky'/>
      <model type='virtio'/>
      <address type='pci' domain='0x0000' bus='0x00' slot='0x03' function='0x0'/>
And now names are more logical:
koos@conway:~$ /sbin/brctl show brdmz
bridge name     bridge id               STP enabled     interfaces
brdmz           8000.4ccc6a8efa4b       no              dmz-minsky

Tags: , ,
2021-10-18 Securing the home network: a separate DMZ network
I have a lot of control over the software that runs on systems at home but there are limits to what I can fix and sometimes things are insecure.

Things like the recent wordpress brute force attacks show that random 'loud' attackers who don't care about the chance of getting noticed will try. I sometimes do worry about the silent and more targeted attackers.

So recently I updated my home network and I now have a DMZ network. At this moment it is a purely virtual network as it doesn't leave the KVM server. Hosts in the DMZ have a default-deny firewall policy to the other inside networks. Specific services on specific hosts have been enabled.

I first moved the development webserver, which allowed me to tune those firewall rules and fix some other errors.

Now other webservers and other servers offering things to the outside world have moved.

Tags: , , ,
2021-07-12 Checking the rcu_sched messages finds repeated mention of cdrom scans
I was going through some rcu_sched messages and noticed kernel routines related to the cdrom drive showed up a few times in the tasks that were 'behind'.
[335894.319961]  [<ffffffffc03d864a>] ? scsi_execute+0x12a/0x1d0 [scsi_mod]
[335894.320702]  [<ffffffffc03da586>] ? scsi_execute_req_flags+0x96/0x100 [scsi_mod]
[335894.321820]  [<ffffffffc04a7703>] ? sr_check_events+0xc3/0x2c0 [sr_mod]
[335894.322551]  [<ffffffffb58224a5>] ? __switch_to_asm+0x35/0x70
[335894.323256]  [<ffffffffb58224b1>] ? __switch_to_asm+0x41/0x70
[335894.323906]  [<ffffffffc047d05a>] ? cdrom_check_events+0x1a/0x30 [cdrom]
[335894.324545]  [<ffffffffc04a8289>] ? sr_block_check_events+0x89/0xe0 [sr_mod]
[335894.325186]  [<ffffffffb551a9a9>] ? disk_check_events+0x69/0x150
Because the virtual machines don't do anything with the virtual cdrom after the first installation I'm removing them from all virtual machines and see what that does for these messages.

Tags: , ,
2021-07-08 Another panic in a virtual machine
At the end of this morning I noticed the root filesystem of the shell server on the homeserver had turned itself read-only. Another DRIVER_TIMEOUT error in the kernel messages. And I didn't want to get to a situation with half of the filesystem in lost+found like the previous time.

This time I decided to use a different approach in the hopes of getting back to a working system faster. And they worked this time.
  1. echo s > /proc/sysrq-trigger to force a sync
  2. echo u > /proc/sysrq-trigger to force an unmount of all filesystems
  3. I killed the virtual machine with virsh destroy (the virtualization equivalent of pulling the plug)
  4. I created a snapshot of the virtual machine disk to make have a state of file system to return to in case of problems in the next steps
  5. I booted the virtual machine and it had indeed filesystem issues
  6. So reboot in maintainance mode and did a filesystem check
  7. After that it booted fine and the filesystem was fine, nothing in lost+found
After things ran ok for a while I removed the snapshot. I also changed the configuration to use virtio disks and not ide emulation. Ide emulation disks have a timeout (DRIVER_TIMEOUT) after which things are given up. The fact that (emulated) I/O hangs for 30 seconds is bad, but maybe related to the rcu_sched messages. Maybe time for some more updates.

Tags: , ,
2021-07-03 Trying a DNSSEC zone signing key (ZSK) rollover
Time to do a zone signing key (ZSK) rollover. That rollover is relatively easy because I don't need to synchronize it with the DS key in the parent zone.

I generated a 'successor' key for and set a short-notice publication date. The old ZSK has keytag 02908 and the new one has keytag 25619. There is an overlap of a month in which both keys are seen as valid because caching of DNS answers mean there can be signatures created with the old ZSK in caches.

Generating a signed zone after the validity of the new ZSK has started shows both ZSKs signed as valid. Old and new zone signing key:
; This is a zone-signing key, keyid 2908, for
; Created: 20190704113915 (Thu Jul  4 13:39:15 2019)
; Publish: 20190704113915 (Thu Jul  4 13:39:15 2019)
; Activate: 20190704113915 (Thu Jul  4 13:39:15 2019)
; Inactive: 20210705000000 (Mon Jul  5 02:00:00 2021)
; Delete: 20210805000000 (Thu Aug  5 02:00:00 2021) IN DNSKEY 256 3 13 lXntnbvQqHy+OSG/2RpHEbcYzeUAB2tFE+d5Us9M07Ndw7TI2DF2TIDx vC3bPomCE2102FJSr8/DnzoRiMHreg==
; This is a zone-signing key, keyid 25619, for
; Created: 20210702115321 (Fri Jul  2 13:53:21 2021)
; Publish: 20210703000000 (Sat Jul  3 02:00:00 2021)
; Activate: 20210705000000 (Mon Jul  5 02:00:00 2021) IN DNSKEY 256 3 13 kJpmrljuP7PncZij7G1Yn9xngKe1xUpuONG2XAx8AYXu//qXClAbgg3B bmzyeDpFAw2gDRhjQ7f5o20c1QK9OA==
So I generated the key on 2 July 2021, with a set publication date of 3 July 2021. I shortened the prepublication period to avoid problems with other things happening in the near future and today it changed to published. If I generate new signatures again on 5 July 2021 those will use the new key.

DNSSEC is a process with lots of things to get your brains around, and a key rollover is one of those things. A key signing key rollover is even harder because uploading of the public key to the registrar has to be kept synchronized with the published information. That is why I am testing all this on where it is not a major problem if something fails.
Read the rest of Trying a DNSSEC zone signing key (ZSK) rollover

Tags: , ,
2021-06-03 Uitgaande mail via xs4all gaat binnenkort niet meer zonder authenticatie
Ik beheer mijn eigen mailserver (met al meer dan 25 jaar sendmail in gebruik) en nu kreeg ik ook de brief over de aanpassingen in SMTP van xs4all. Het komt er op neer dat relaying op basis van IP adres gaat verdwijnen.

Om een helpdeskramp te voorkomen gaat het uitschakelen per gebruiker. Ik heb een brief gekregen dat ik soms gebruik maak van deze route en dat moet aanpassen.

Dat klopt, voor sommige servers was het feit dat ik weinig mail naar die servers stuur een reden om het te blokkeren. Of het ooit ontbreken van een IPv6 reverse pointer. Dat laatste heb ik goed laten zetten toen.

Op de website van xs4all staat wel een uitleg: Veilig e-mailen 2020 - xs4all maar daar staat niets bij over sendmail. Thuisservers die mailen zijn blijkbaar niet meer hun doelgroep (mijn Cron Daemon is er anders best goed in!).

Ik ben maar eens begonnen met het leeggooien van de lijst in de mailertable. We gaan zien welke domeinen nu onbereikbaar zijn.

Tags: ,

IPv6 check

Running test...
, reachable as PGP encrypted e-mail preferred. PGP key 5BA9 368B E6F3 34E4 local copy PGP key 5BA9 368B E6F3 34E4 via keyservers

Meningen zijn die van mezelf, wat ik schrijf is beschermd door auteursrecht. Sommige publicaties bevatten een expliciete vermelding dat ze ongevraagd gedeeld mogen worden.
My opinions are my own, what I write is protected by copyrights. Some publications contain an explicit license statement which allows sharing without asking permission.
Other webprojects: Camp Wireless, wireless Internet access at campsites, The Virtual Bookcase, book reviews
This page generated by $Id: newstag.cgi,v 1.37 2022/02/15 21:48:19 koos Exp $ in 0.026125 seconds.