We have just received an information from our datacenter operator, Softlayer, about their backbone network upgrade on Sunday, 26. Sep at 6.00 - 10.00 AM GMT. There might be temporary network issues that could affect availability of Wikidot, so please be aware of it.
Here is the original message for more details:
SoftLayer Scheduled Maintenance - Short Notice [09/26/2010]
Date: Sunday, September 26, 2010 (09/26/2010)
Start time (GMT+00:00): 06:00 AM
End time (GMT+00:00): 10:00 AM
Services affected: Public Network / Backbone
Location: WDC01
Duration: 4 hours
- Short Notice Maintenance *
SoftLayer Engineers will be conducting maintenance on the transport network that provides connectivity between the WDC01 datacenter and the Equinix facility in Ashburn. During this window, Engineers will be migrating a number of transport circuits from existing gear to new dark fiber gear. Once complete, this new system will allow SoftLayer to scale capacity between these sites from the existing 60 Gbps limit to well over 800 Gbps if required.
The following services will be impacted during this maintenance:
- Public network connectivity for the WDC01 facility
- Public / Private connectivity to other SoftLayer facilities from the WDC01 facility
During this maintenance, Engineers will make every effort to mitigate the impact to customer traffic; however, customers are likely to see routing anomalies, packet-loss, and increased latency.
Softlayer engeneers have proven more than once they are real experts in their field, so I believe the upgrade will be smooth and without major issues.
Update
As far as we know, there were not reported problems with Wikidot during the maintenance window. However we experienced some problems afterwards, i.e. on Monday night from 0:00 AM to 1:15 AM GMT. The problems are resolved now and our network is working well.
Perhaps I'm off base on this, but our site did experience connectivity issues starting around 830pm (-05GMT). Reading the above, this sounds like it was routine (scheduled) maintenance to enhance the capacity of the network/farm. If thats the case, I think more than a few hours notice is not unreasonable.
In addition to more than a few hours notice, I am of the mindset that more than a blog post should go out when the potential for a system down event is there. Perhaps a distribution list to send notification out that something like this will occur. Given that there were some connectivity issues, how can I be expected to read a blog post on the site thats not working? Its like sending am email to let everyone know the email server is down? :p
In fairness to Michal and Wikidot, reading the above there are 2 key points you seem to have overlooked:
The notice from SoftLayer says "Short Notice Maintenance" and Michal said "We have just received an information from our datacenter operator, Softlayer…".
The notice was more than a few hours - the blog post is dated 23 September. I think you make a good point about perhaps looking at a different method to alert us to this potential downtime, but I believe Michal made an effort to provide this information in a timely manner and passed on the information as soon as he heard about it.
Community Admin
Not sure how this can be managed if the time would be really "short" - I am used to have a look on twitter:
http://twitter.com/wikidot
to see what wikidot is saying here..
Service is my success. My webtips:www.blender.org (Open source), Wikidot-Handbook.
Sie können fragen und mitwirken in der deutschsprachigen » User-Gemeinschaft für WikidotNutzer oder
im deutschen » Wikidot Handbuch ?