The brief web outage today lasting approximately two minutes was to apply a security update to the Apache server bringing it up to 2.4.62 owing to vulnerabilities found in the previous version.
At the same time, I also upgraded the kernel to 6.10.0. The 6.10 kernel has some improvements that speed up encryption.
Some update pushed on Rocky8 and Fedora broke rwho and ruptime on those machines. They will still provide user status to other servers but are no longer pulling other servers for rwhod.
Further, ruptime requires the ‘daemon’ command which is no longer available on these machines.
At some point NIS will disappear from Fedora and when it does we will be forced to retire the machine at that time. Because we don’t know when this will happen we can not predict this date. Therefore we recommend NOT relying on that machine for anything. If you have cron jobs on it, please move them to rocky8 or if you do not need a Redhat environment, one of the other servers.
Roundcube mail is back. It is in the WebApps->Mail->Roundcube menu or the direct path is https://roundcube.eskimo.com/.
This is running on the new server and will be down between Wednesday evening and possibly Saturday though I may be able to get it back up sooner. I am replacing the motherboard, AGAIN, as the new motherboard is stable BUT has two dead memory slots so the machine is only seeing 192GB of 256GB of installed memory.
None the less, it is now an 18 core machine, 36 threads, at 4.8Ghz all cores.
The way I finally got roundcube working is that I installed Ubuntu 20.04 on a virtual machine on the new server. I have this virtual machine listening to a non-routeable IP address so that it can not be reached directly from the outside world to provide some degree of security to ancient software. Then I have it proxied through the main web server on the new machine which listens to both a public IP address and a private one that it can use to talk to internal servers. Something about 24.04 breaks roundcube, I know not what.
As previously noted, centos7, scientific7, and now centos-stream are going away. Centos7 and scientific7 at the end of this month, and centos-stream in a few hours as Redhat has already dropped the repositories for it ahead of schedule.
I’ve brought two new distros online, rocky8.eskimo.com is primarily a replacement for centos-stream, centos7, and scientific7. It is much like Centos WAS before Redhat took it over. Please let me know if there are applications that are presently missing or not working.
The other new shell server I brought online is popos.eskimo.com. Popos is a disguised Ubuntu (they are like Zorin, a Ubuntu with some of their own stuff placed on top). Because of the popularity of the release with the Linux community at large and the fact that Ubuntu based systems are relatively easy to maintain, I’ve added it.
If you have any other distros or services you would like to see here please contact nanook@eskimo.com.
Mail lists and some aliases are currently out of service owing to the aliases file on the client mail server somehow being truncated and half not there. I am restoring from backups. While this is in progress mail lists will not work as ALL aliases for smartlist are gone and many users aliases are gone at the moment.
(Update the aliases file is restored, lists are back in service)
On other news, while waiting for the new CPU chip to get here I had the machine sitting in BIOS configuration screen and after about three days it shut itself down spontaneously.
Since I had already replaced the motherboard (some bent pins on the CPU socket necessitated this), I went ahead and replaced the power supply. Prior to replacing it, it would not run at all above 4.3Ghz and would not run stable even at 4Ghz, now it is running at 4.7Ghz, so there was some issue with the old supply, either a ripple problem or it would just not capable of supplying enough power on the 12v line. Since this CPU can use as much as 540 watts it requires a very robust supply.
I have the new server operational with the new motherboard and OLD CPU. However, it will not operate at rated speed let alone overclock. There is something wrong with the memory controller and it will not operate at the rated 2933 Mhz under Linux, oddly it will run at over 4300Mhz using memtest-86 so there is something about the way Linux accesses memory that is different than memtest-86 and I know not what. I suspect it has to do with system mode where the address is direct verses user mode where the address is translated.
However it is running at lower speed with the new motherboard so that is better than not at all. I have a new CPU on order, the first one was dead on arrival, the motherboard won’t even recognize it as being installed. I’ve got a bit more work to do on it, I have to upgrade the firmware and UEFI bios on this motherboard, and then if the new CPU isn’t here yet I’ll return it to service at slightly below normal speed until the new CPU arrives.
The new server continues to be unstable, it tends to spontaneously reboot when the upper reaches of memory get used. It may be a bad DIMM. I also accidentally over-volted the CPU at one point, I meant to set 1.17 volts and accidentally set 1.7, the max for this CPU is supposed to be 1.5v so may have broken down a transistor or two.
When I bought the RAM for this machine I bought 8x32GB DIMMS as 64GB DIMMS were unavailable at the time. The Intel engineer did warn me that the CPU had difficulty driving two DIMMS per channel and I probably would not be able to run memory at the memory controllers rated speed, none the less until recently it was running OK.
Since 64GB DIMMS are now available, I’ve ordered 4x64GB to replace the existing 8x32GB, that way the CPU only has to drive one DIMM per channel and can operate at maximum speed. I also got a steal of a deal on an i9-10980xe, this is basically an 18 core version of the i9-10900x. So when this all arrives I will be replacing all the RAM and the CPU. Also, I purchased a 3000 RPM max CPU fan to replace the 1500RPM fan on the Noctua 15 cooler, with 18 cores I will need all the cooling I can get. Though this CPU really does not get real hot until you clock the cores at 4.3Ghz or above, below that you can keep the voltage down to where it stays reasonably cool even under load.
The issues that remained with Centos7 and Scientific7 with the existing 6.0.x kernels were not expected to be fixed until 6.2 according to the developers, however, 6.0.12 appears to have fixed them as both machines booted cleanly after the upgrade last night.
Also, I adjusted the clock on ice from 4.3 Ghz to 4.2 Ghz as it was ever so slightly unstable at 4.3 Ghz.
Lastly, Manjaro is restored, but your existing ssh keys will not work as this is a fresh install so you will need to delete and create / propagate new keys. There is much software not yet installed, so if something you need is missing please send a note to support@eskimo.com and I will prioritize.