backend corruption

[slimjen]

Hi All. I hope someone can assist me with this problem. If I’m not in the right area; please point me in the right direction. I’ve had this database created in Access 2010 for about 2yrs. I do have it split with about 100 frontend users. The backend is on a shared drive. About two weeks ago; I came in to a mirage of emails that the users were getting an error “unrecognizable database…….” I immediately realized the backend had to be corrupted. I went to the shared drive and the backend said 0kb. I had to have our IT dept. restore it because I don’t usually have a reason to back it up locally. So; after it was restored; it happened again about mid-day. By this time, I was looking for the culprit. I started with the IT dept. to see whether they pushed some kind of update that didn’t agree or something. They insisted they did nothing. Something changed but I couldn’t find out what. To make a long story short; I did find out that now some of the users have been upgraded to Access 2016. So now I have some users on 2010 and some on 2016. This is normally no big deal because the database is developed in 2010. So I am scratching my head because nearly every day; the backend gets corrupted. I’ve tried several things including reducing the number of users using the database at one time and making copy and repair every couple of hours. I cannot figure out the problem and our IT dept. is no help. Can anyone possibly has had the same issue or anyone can shed some light on this situation please.

[rpeare]

does your database run a compact/repair on exit? I've found even between say access 2013 and access 2007 (which are supposed to be compatible) there are enough subtle differences that it can make the database unreadable to the lower version user and for me it was related to the compact/repair function I had to disable it on all of the higher version users.

[slimjen]

No. I learned this a long time ago that it wasn't wise at all to have this setup automatically. Thanks

[rpeare]

Are you using multi value fields or calculated fields in any of your tables?

Can you identify a specific process or processes that are bombing out?

[slimjen]

No calculations in the tables. I went through all tables and queries to see but it's taking such a long time in such a large database.

[aytee111]

Have you tried recreating the database? Just throwing this into the mix, not sure if it would help. The issue might have occurred the first time and now is being carried on.

[slimjen]

Will try this when I get the chance. With over 100 users in the database constantly; I didn't want to take the chance losing it completely. thanks

[NTC]

back end corruption is very rare in a correctly deployed split database. It is very common when incorrectly multiple users are directly in the file at the same time (not via their front end). The only other cause I typically see is when the server itself has a hiccup like a lightening strike or something like that.

when off line I would suggest you create a brand new database, and then import all the tables. the back end should have nothing but tables.

[rpeare]

Are your forms bound forms or unbound forms? if you are splitting the database and each user has their own front end and they are all accessing the same table at the same time using bound controls this could cause a problem as well, I don't know why it would show up now after 2 years of not having a problem, it's just extremely suspicious to me it started occurring after a version update. Have there been any modifications made to the database from a computer with 2016 on it? if you are the administrator that would be your machine. If you have I would try to revert to the 2010 version (if you still have it prior to you upgrading) and only made design changes/modifications from a 2010 version machine.

[redpenner]

We were having a lot of BE corruption problems (i.e., daily) a couple of years ago after a SERVER update, and found that taking both the BE and the FE's off the server resolved the problem. (Staff had all of their files except software on the server, even "My Documents"). The FE's now reside on staff hard drives and the BE resides on my hard drive. The permissions needed aren't that hard to set up and maintain and the connections have never failed, and not a single episode of corruption. I just can't turn off my computer...

[NTC]

It would make me wonder if the server had a feature to shadow back & forth between drives for redundancy - and that feature was screwing up the indexes if occurring while the database was live in operation.....