Grid redundancy failiure

At 14:50 a PDU in HY DC failed.
This caused some equipment to lose grid redundancy.
At 20:30 , all equipment was moved to temporary failover grid redundancy until DC operator can remedy failed PDU.
The root cause was a short circuit in a server power supply causing fuse to trip.
On Tuesday during office hours, DC operator will attempt to remedy the tripped fuse and move equipment back to original PDU.
Impact: No service disruption of network connectivity (redundant power). Reduced Grid redundancy between 14:50 – 20:30.

Co-location electricity prices (updated with clarification)

Hello,

It has most likely not gone unnoticed by anyone that the electricity
prices are hitting record highs across Europe. During the winter
months, we know from experience that the price of electricity goes
up, something we always take in consideration when we price our
services, but this year we’ll need to adjust our electricity price
for those of you who have colocation, rent an entire rack or have dedicated servers.

Last year the price of electricity increased by over 500 percent in
Sweden, compared to 2020. The new price level has continued into
2022 and we see no end to these electricity prices in a foreseeable future.

The adjustment of the electricity price will be activated in the
months where the average electricity spot price exceeds SEK 0.75/kWh.
On these occasions, we will adjust the electricity price by SEK 10
per kW per month for each Swedish öre it exceeds SEK 0.75/kWh.

For customers in Stockholm and Falkenberg, the spot price will be
regulated according to electricity area SE3 resp SE4.

Price example: if the average electricity spot price during a month
is SEK
0.85 / kWh, we will charge SEK 100 extra per kW and month.

The price adjustment will take effect first of March this year!

This electricity price change will go in effect on agreement renewals for all co-location and/or dedicated server customers who have been notified, this this new electric price calculation will complement any previous included electricity and be based on known or calculated electric usage of customer.
This is to safeguard Adminor against the exceptional price swing on electric market which is plaguing Europe at this time. Adminors pricing is transparent and matches the electric price of our provider and aims to not make any additional money from the electric price billing. If our provider changes their price calculation we will adjust this as well, this may mean that your price will change quickly. We’ve taken all steps we can to offer reliable electricity , and alternative hosting sites – this is offered at first come, first served basis.

Utbyte switch

Vid 23:45 17/2 gick ett linjekort på en switch sönder. Switchen stängdes av. Inga tjänster påverkades då switchen är speglad.
04:30 18/2 påbörjades utbyte av switch, vi passade på att uppgradera till ny infrastruktur i noden där switch byttes ut.

Detta krävde kortade nedtid för vissa kunder för nätverksförändringar. Några få kunder fick avbrott som var längre men dock inom SLA-villkor.
Vid 06:30 var allt arbete slutfört och noden har uppgraderats till ny generation av switchar som medför kraftigt ökad kapacitet.

Planned maintenance work 2021-06-20

When: Sunday 12.00 – 18.00 on 2021-06-20
Why: Datacenter facility provider maintenance
Affects: Hammarby Datacenter, Interconnect between Hammarby DC and other datacenters
as well as other services on 46.253.192.0/20 hosting range

Details:
Our datacenter provider GlobalConnect are performing structural integrity improvements in the facility. This requires us to move equipment in the zone we’re located in. We’ll be doing this with ”construction electricity” to avoid powering down the equipment.

We need to move some uplink cables, which may cause momentary interruptions.
Our intention is that the disruptions will only last minutes at a time, but we’re still announcing a service window. Please plan your activities and services according this this service window.

Network issue 2021-05-09

We had an issue with a core switch for our bladechassis that did not failover as it should following a fault.
Switch was still reporting functional and secondary switches did not take over as they should. This caused delay in troubleshooting process.

Effected datacenters: VBDC , interconnection between VBDC and HY
When: 12.55 CET – 13.41 CET
Component: dr-se-vb-sw02

We’ll perform software updates during night time to prevent any software causes if this issue.
If it still happens again, we’ll replace the switch as faulty.

Maintenance window / router maintenance

Adminor is scheduling a router maintenance effecting customers on IPranges: 46.253.192.0/20 and 185.111.240.0/22 .

Primary Service window is 2021-03-22 Monday 00.00 – 04.00 (two week notice!).

We are aiming for an outage less than 30min (15min or less usually), but keeping the window longer incase we hit an unforeseen snag or hardware issues that needs to be resolved.

Emergency service window (if we need to schedule sooner) 2021-03-15 Monday 00.00 – 04.00 CET.

We will try to keep this maintenance brief but are letting you know as our valued customer.
 
Best regards
Adminor AB

Azure and Azure files MMAP broken

We found out (or rather re-discovered) that Azure Files CIFS implementation does not correctly support mmap function that is used in some softwares.

It has been previously reported on https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=900821 but we thought it might be worth clarifying this for people spending a lot of time googling why their apache2 fails to deliver files larger than 1MB.

The default setting in apache2 does not mmap files smaller than 1MB hence why the issue is not noticed on small files or files not hosted on the CIFS volume. Large files from the affected volume will however cause a ”Read error at byte X/ X (Bad address). Retrying.” error if you try to wget or download the files via apache2 on a web app using Azure Files mounted CIFS storage (path mapping).

Some searches might lead you to investigate apache2 file limitations preventing files larger than 1MB from beint transmitted.
LimitRequestBody is not a fix for this issue.

Scenario:
Azure web app or VM with Apache2 or other mmap using software fails to read files properly from Azure Files mounted ”path mapping” volumes which uses CIFS.

Workaround:
EnableMMAP off

This configuration line can be done per affected Directory, global apache2 config or in the vhost.