discussion Important Update: New Sysadmin + Server Move Plan

[hawaii]

Everything will be good . I trust in you !!!

 

[NotLooking]

Days after saying only two people hold the keys to the kingdom here as a reason to have faith in the site the server loses its boot partition and the admin goes MIA... "Everything is fine"

 

[Jack Of Blades]

Days after saying only two people hold the keys to the kingdom here as a reason to have faith in the site the server loses its boot partition and the admin goes MIA... "Everything is fine"

dev goes MIA because of the war, and the new dev is trusted.

 

[LewdSardin]

When is this move going to start? Just so I know when not to visit

 

[Jack Of Blades]

When is this move going to start? Just so I know when not to visit

sysadmin is trying to move the server with rsync so no need to put the site offline/closed.

This post has been automatically merged by red99 legendary mod. Wednesday at 9:44 AM

When is this move going to start? Just so I know when not to visit

You must be registered to see attachments

 

[Nazha_Aelius]

How long does it take to move a server, if all goes well?

 

[Jack Of Blades]

How long does it take to move a server, if all goes well?

honestly depends on how slow the system is.

 

[lonestar69]

but could it be hours or days? Sorry, I'm ignornant of how long these things might take

 

[sharkz]

but could it be hours or days? Sorry, I'm ignornant of how long these things might take

I'm guessing that depends on the amount of Data and how much dataspace clogged up over the past years, i ain't in no way a Dev of any sort but i sure can imagine at least a couple days
even if everything goes according to plan but in the end it's impossible to tell i'd say.

 

[VSCDB]

I'm guessing that depends on the amount of Data and how much dataspace clogged up over the past years, i ain't in no way a Dev of any sort but i sure can imagine at least a couple days
even if everything goes according to plan but in the end it's impossible to tell i'd say.

Basically. Even tho "clogged up dataspace" is quite funny to hear when you do have some CS knowledge. More modern storage does not really "clog up" but in the past there was disk fragmentation, which was 1. Mostly a windows problem and 2. Isn't really an issue concerning users anymore. It still can happen, only with HDDs tho, and is basically files being fragmented over the whole disk and therefore need to be read from multiple, sometimes further sometimes closer, physical locations, which makes reading very slow. This is usually not necessary anymore tho since I think Win 7 or 8 they automatically defrag the system, and Mac and Linux have a two different technologies to prevent/minimize this. In Linux it can happen that you need to manually defrag, but usually is not necessary because of how ext4s filesystem works.

To the original question.
Basically depends on a couple factors, first, are they on the same physical premises, how much data is to be transferred, how its being transferred and from what kind of drive and to what drive is being transferred.
So basically will take longer when they are in different physical locations since you either have to go via an internet connection (most likely here) which will be limited, at max usually at around 1Gib/s could be a bit more, unlikely you even can use the full bandwidth, and might be less here as well. Then the actual storage devices, a good ssd with fast and big flash will shorten the time quite a bit, an HDD or cheap ssd will likely take quite a bit longer. The method used is also pretty important, if as previously mentioned, rsync is used, its basically single threaded, meaning everything will be done step by step, while the OS could optimize a bit of overhead (often times a lot, but here file I/O will likely restrict it) to make it more efficient if used multi threaded. There are a couple more variables, but since I already wrote too much, and these are the most important, I'll end it here.

TL;DR: anything from a couple hours to days is possible, depending on the infrastructure and tooling.

This post has been automatically merged by red99 legendary mod. Wednesday at 1:49 PM

Would say likely, from a technical standpoint, without knowing details, probably not longer than 12h to a day. But well can't say for sure.

 

[sijas]

Basically. Even tho "clogged up dataspace" is quite funny to hear when you do have some CS knowledge. More modern storage does not really "clog up" but in the past there was disk fragmentation, which was 1. Mostly a windows problem and 2. Isn't really an issue concerning users anymore. It still can happen, only with HDDs tho, and is basically files being fragmented over the whole disk and therefore need to be read from multiple, sometimes further sometimes closer, physical locations, which makes reading very slow. This is usually not necessary anymore tho since I think Win 7 or 8 they automatically defrag the system, and Mac and Linux have a two different technologies to prevent/minimize this. In Linux it can happen that you need to manually defrag, but usually is not necessary because of how ext4s filesystem works.

To the original question.
Basically depends on a couple factors, first, are they on the same physical premises, how much data is to be transferred, how its being transferred and from what kind of drive and to what drive is being transferred.
So basically will take longer when they are in different physical locations since you either have to go via an internet connection (most likely here) which will be limited, at max usually at around 1Gib/s could be a bit more, unlikely you even can use the full bandwidth, and might be less here as well. Then the actual storage devices, a good ssd with fast and big flash will shorten the time quite a bit, an HDD or cheap ssd will likely take quite a bit longer. The method used is also pretty important, if as previously mentioned, rsync is used, its basically single threaded, meaning everything will be done step by step, while the OS could optimize a bit of overhead (often times a lot, but here file I/O will likely restrict it) to make it more efficient if used multi threaded. There are a couple more variables, but since I already wrote too much, and these are the most important, I'll end it here.

TL;DR: anything from a couple hours to days is possible, depending on the infrastructure and tooling.

This post has been automatically merged by red99 legendary mod. Wednesday at 1:49 PM

Would say likely, from a technical standpoint, without knowing details, probably not longer than 12h to a day. But well can't say for sure.

there's really no need to limit that, could be two weeks if you have enough data
honestly, i think none of us actually know
- how much data there is
- how thin the line is
- how strong the machines on either side are

It can literally be any number, so I suppose this is more of a "wait and see" scenario

 

[VSCDB]

there's really no need to limit that, could be two weeks if you have enough data
honestly, i think none of us actually know
- how much data there is
- how thin the line is
- how strong the machines on either side are

It can literally be any number, so I suppose this is more of a "wait and see" scenario

For sure, but its unlikely they have more than a couple tens of terabyte since AFAIK, they are not actually hosting most files (some videos and pictures like pfps). And source code, usually, doesn't take that much space, and I hope if they use precompiled languages, that they won't rebuild on that machine but have a prebuilt packages or VMs/Containers.

But yes, that's why I said that probably it shouldn't take longer but not its guaranteed it won't. It could take 2 weeks, but then I have to ask if they have a full file storing service running as well, or how bad the connection between these machines is. I come from the scene, had to migrate multiple servers and clusters already.

All of this anyway is dependent on when the actual transferring happens, since we don't know if that even already started.

This post has been automatically merged by red99 legendary mod. Wednesday at 2:45 PM

BTW just to make sure, not here to criticize, hurry or attack anyone, just wanted to offer insights from my profession. I know how much work this is and respect the new sys admin for actually taking up the job and the team for making these decisions.
Sometimes I am not the best in writing

 

[sharkz]

BTW just to make sure, not here to criticize, hurry or attack anyone, just wanted to offer insights from my profession. I know how much work this is and respect the new sys admin for actually taking up the job and the team for making these decisions.
Sometimes I am not the best in writing

But was very informative and have a slightly better understanding on how it's being processed.

 

[ironbuilt]

Good luck with the move

 

[VSCDB]

But was very informative and have a slightly better understanding on how it's being processed.

Then I am always happy to help

 

[cyberpunk667]

Just as long as the shop get fixed been waiting to redeem my Donation Coins for members+

 

[Republican Bear]

Damn shame about whats happening with Apollo. Hope he's fine and communications are not down as reported on the news.

Hope the new server works better and all errors get ironed out before you go live once again as well.

 

[KTrash]

Is there no alternative to Discord? It was already hit or miss with people, and now with the age verification shit theres just no chance of using it now.

there are decent alternatives, matrix for one is really good.

i am working on an alternative, but it isnt finished yet, i still need to code in E2EE

and matrix has e2ee? especially decent when turning off federations, also adding a simple script that does history purges every now and then to minimise disk usage and other things.

 

[Xanxan6969]

All good stuff, peace out

 

[Xeron_Xr]

praying for apollo to make through.