Under EPDM I have a doubt about an option in the replication schema configuration:
See attached document.
Here is an explanation of the replication setting called " DAILY REPLICATION "
- Folder to be replicated : C:\EPDM-DUBUS-GROUP
Replicate news or update data from SERVEUR5 (Source Server - Site "B") to SERV-EPDM (Target Server - Site "A")
Start time 8:30 p.m. Monday to Thursday.
- Folder to be replicated : C:\EPDM-DUBUS-GROUP
Replicate news or update data from the SERV-EPDM (Source Server - Site "A") to SERVEUR5 (Target Server - Site "B")
Start time 11:30 p.m. Monday to Thursday.
Where I have a doubt is on the advanced option: "Stop replication after:" 7h and 00 mn
How EPDM behaves:
- it starts replication from site "B" at 8:30 p.m. and stops at 3:30 a.m., simultaneously it starts replication from site "A" at 11:30 p.m. and stops at 6:30 a.m.
- it starts replication from site "B" at 8:30 p.m. and it starts replication from site "A" at 11:30 p.m.: whatever happens 7 hours later (from 8:30 p.m., the time of the replication schema start), it stops all replication at 3:30 a.m.
schema_de_replication-2.png
Be careful, I give this answer but I leave it to others to confirm (or not) my words.
As a general rule, the tempo in a base works via a timestamp system (I said as a general rule).
This normally gives the following diagram:
- At 8:30 p.m., replication of the first site until 3:30 a.m.
- At 11:30 p.m., replication of the second site until 6:30 a.m.
A base works by stain so even if the pattern is the same, the stains are different. The tempo should therefore be applied to each task and not to the entire scheme.
Hoping I have been clear.
1 Like
Hello
Normally it's scenario 1 as jose says
Is there anything that makes you doubt it?
1 Like
Computer-wise, I would say the same ^me thing as colleagues.
However I am doing an EPDM admin training today and tomorrow, I will try to give you an answer until then.
1 Like
Yes Benoit, I have a doubt, (and I put the doubt to the technical support !!!) we were forced to stop the replication (one day when I was not there so I lost the basic setting !!!) following the "crash" of 2 hard drives on the "B" site.
1 Like
Yes, I've heard about it... that's 3
Did it happen when you started a full replication from server A to B? Or was it a partial copy?
If it's a full copy, you may have a huge file that didn't have time to be transferred in 1h30... It's already been seen...
Ask @fifounet44 to free up the internet line and everything will be better... Huh my Fifi?
Let's be serious: do a test at noon for a few minutes instead of several hours, making sure that site B has produced files...
I need to stop replication until site "B" receives the hard drive with all the EPDM data hard from site "A" (today or tomorrow).
Indeed, when we put the disks back on the "B" site, the RAID5 system tried to repair the data. He repaired part of the 330GB and/or recovered part (270GB).
In these 270GB there are some files (difficult to identify) that are corrupted.
(FYI in the weekend via our internet connection we only managed to pass 16GB !!!)
From this simple observation, we decided to send a hard disk with the 330GB, Replace "hard" all the data of site "B" and resume replication only once the data was copied.