| Author |
Replies: 90 / Views: 8,040 |
|
|
|
Bedrock Of The Community
United States
10597 Posts |
|
|
They didn't know when I checked last week. But reloading everything must be time consuming under ordinary circumstances. The way things are right now........? |
Send note to Staff
|
|
|
Pillar Of The Community
United States
2941 Posts |
|
|
Quote: Actually the hard drive got fried, according to my info. RAID arrays are your friends, folks. |
Send note to Staff
|
|
|
|
Pillar Of The Community
United States
6430 Posts |
|
|
Especially for web servers. RAID 10, and at the very least RAID 1 at the minimum. And regardless, there should be nightly off-server backups. It should be no longer than a 1-day restore process at the longest. Datacenter staff should be able to handle it relatively quickly.
Unless they are hosting everything onsite at the PF offices and access due to social distancing is the problem. |
Send note to Staff
|
|
|
|
Moderator

United States
12330 Posts |
|
|
I agree with Dan.
After losing data twice while using RAID, I gave up relying upon RAID. (For example, what do you do when you when a controller fails and sporadically writes crap to all drives for a few hours?)
For a situation like PF, which does not really need instant real-time data protection 24 hours per day, using drive imaging backups would be by far the easiest and most inexpensive solution. Sector level drive replication is blazing fast to do and gives you a perfect duplicate of the entire drive. Then simply change out the imaged backup drives once a day and you end up with a nice weekly backup.
This setup does not take someone with years of IT experience to implement nor does it take someone with experience to rotate the drives or switch over to one of the backup drives.
It also prevents the scary situation of performing a non-reversible RAID configuration procedure of switching over to the secondary drive after the primary drive has failed. This procedure is rarely used and the screens unfamiliar, you sit there staring at the 'less then helpful' screen and knowing that you have to be 100% sure that you are reassigning the drive destinations correctly. One misstep you lose your last remaining drive and all the data on it.
But with a set of replicated fully imaged drives always available, it takes less than a minute or two to retrieve a single file, a directory, or replace an entire drive if it fails.
And if desired, you can setup an even more robust backup process by extra adding drives and removing some from the rotation. These extra drives can then be removed from the building. RAID does not help you if the entire server catches fire or the location is hit with a tornado.
I also use drive image replication on external drives here at home with my computers around the house. Fast, easy, inexpensive and very reliable. Don |
Send note to Staff
|
|
|
Valued Member
Norway
450 Posts |
|
|
Bedrock Of The Community
12553 Posts |
|
|
I saw your comment Lorry and checked for myself. It is pretty shocking that it is still down. It will be a month soon. Wow. |
Send note to Staff
|
|
|
Pillar Of The Community
United States
6430 Posts |
|
|
Expertizing organizations in philately are a shadow of what exists in numismatics. Even before I got out of coin collecting approximately a decade ago, if I sent in coins to NGC or PCGS, the two premier certification companies, I could log into my account on their websites, see the statuses of my submissions, get notifications when grading was done, etc.
Philatelic expertizing organizations are operating using antiquated analog methods. They are a full 20 years behind the times. No email notifications when submissions are received or completed. No tracking provided when shipments are sent. No accepting credit cards or PayPal, only checks. No end user account functionality on the websites.
It's an industry operated by old people for old people using old mindsets.
I'm sorry if that sounds harsh, but as capable as their philatelic knowledge may be, their adoption of technology in the daily business aspects is atrociously poor.
This site should have been down for at MOST a week, and that's giving them far too much latitude. Assuming they weren't running the site on an 80386 in the back room or on a Go Daddy $10/month shared hosting plan, and actually have a business-level hosting arrangement and/or IT functionality, this should have been resolved within days, even taking the coronavirus into consideration. If they didn't have some sort of cloud-based backup with nightly incrementals and weekly snapshots (at the worst), someone should be relived of IT duties pronto. |
Send note to Staff
|
|
|
|
Moderator

United States
12330 Posts |
|
|
Quote: ...They are a full 20 years behind the times... Well, in their defense, they are in stride with the rest of the hobby.  Early adoption would have been around 1996-8, a bit late to the game would have been around 1999-2005, really late would have been in the 2006-2011 time frame. Not sure what to call it now... Don |
Send note to Staff
|
|
|
Bedrock Of The Community
12553 Posts |
|
|
Pillar Of The Community

United States
1816 Posts |
|
|
That's good news. I sent in a package just a few days ago. Given the circumstances I am glad that 1) someone was there to sign for it and 2) they acknowledged receiving it in email. They may not be up on tech, but at least they are up and running, and according to their recent mail - all healthy. |
Send note to Staff
|
|
|
Pillar Of The Community
United States
3485 Posts |
|
|
Pillar Of The Community
3859 Posts |
|
|
Are there any other expertizing services that also have searches of their certificates available online? |
Send note to Staff
|
|
|
Moderator

United States
12330 Posts |
|
|
Bedrock Of The Community
United States
10597 Posts |
|
|
Moderator

United States
12330 Posts |
|
|
That kind of feedback makes it feel the all the effort that was put it worthless; if it has such serious limitations I guess I should take it down. Don |
Send note to Staff
|
|
Replies: 90 / Views: 8,040 |
|