Providence_Healthcare
asked on
Exchange Replication Failed, Out of Log space
Hi,
I've inherited an Exchange 2007 cluster running on W2k8. There are 2 machines in the cluster - EX01 and EX02. The cluster name is EXCLUS. EX01 is typically the active host.
A while back, we the Server service on EX01 failed (with service specific error 2182 (0x886)). Since this time, backups have not been running correctly. As a result, we exhausted all log space.
As a short term solution, I moved some of the log files to a separate drive.
If I start the Server service on EX01, will replication start again correctly? Should I anticipate any problems?
Common Event ID's on EX02: Source: MSExchangeRepl - Event ID 2074 (this is directly attributable to the non-functional Server service on EX01).
Thanks
I've inherited an Exchange 2007 cluster running on W2k8. There are 2 machines in the cluster - EX01 and EX02. The cluster name is EXCLUS. EX01 is typically the active host.
A while back, we the Server service on EX01 failed (with service specific error 2182 (0x886)). Since this time, backups have not been running correctly. As a result, we exhausted all log space.
As a short term solution, I moved some of the log files to a separate drive.
If I start the Server service on EX01, will replication start again correctly? Should I anticipate any problems?
Common Event ID's on EX02: Source: MSExchangeRepl - Event ID 2074 (this is directly attributable to the non-functional Server service on EX01).
Thanks
ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
If I start the Server service on EX01, will replication start again correctly? Should I anticipate any problems? - All DB's will go into Failed state.
Once you move the Logs you will have to do a Reseed :(
- Rancy
Once you move the Logs you will have to do a Reseed :(
- Rancy
ASKER
The command format - would it be something like:
Suspend-StorageGroupCopy -identity "User Storage Group"
(all other options seem optional).
and:
Resume-StorageGroupCopy -identity "User Storage Group"
(all other options seem optional).
Suspend-StorageGroupCopy -identity "User Storage Group"
(all other options seem optional).
and:
Resume-StorageGroupCopy -identity "User Storage Group"
(all other options seem optional).
If your moving logs out or to another location you will have to Reseed ...
Suspend and then resume works if your doing a Maintenance of restart of Passive or some small activities and not touching DB or Logs.
- Rancy
Suspend and then resume works if your doing a Maintenance of restart of Passive or some small activities and not touching DB or Logs.
- Rancy
ASKER
Rancy,
Yesterday while Exchange was down (due to lack of log space), I set several of our stores to circular logging.
Will this affect my actions?
(I normally don't have to deal w/ Exchange, I'm not very experienced with it).
Yesterday while Exchange was down (due to lack of log space), I set several of our stores to circular logging.
Will this affect my actions?
(I normally don't have to deal w/ Exchange, I'm not very experienced with it).
This happened to me recently where one of DB drive (DB9) was full. This was due the weekly full backup and daily backup on run on this DB was failed. I'm running on Exchange 2007 CCR.
The drive of DB 9 was full of log and this cause DB9 dismounted. So what I did is:
1. Verify if DB9 is clean shutdown.
2. If yes, then manually delete the log files and mount the DB back
3. Using EMC, and verify DB9 copy status.
4. If copy status failed then at passive node, click on Update Storage Group Copy
5. This will delete the DB and log file on passive node and start synchronize again.
The drive of DB 9 was full of log and this cause DB9 dismounted. So what I did is:
1. Verify if DB9 is clean shutdown.
2. If yes, then manually delete the log files and mount the DB back
3. Using EMC, and verify DB9 copy status.
4. If copy status failed then at passive node, click on Update Storage Group Copy
5. This will delete the DB and log file on passive node and start synchronize again.
SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
Yesterday while Exchange was down (due to lack of log space), I set several of our stores to circular logging.
Will this affect my actions? - Yes sir reason being the Database are now in Split brain situation so only way out of it is Reseed those Databases to Passive node.
Yeah if Backup is taken today with Replication broken for some DB for whatever reason once i fix and reseed it will purge logs as its criteria of "Copy and replay to passive DB" is met :)
Lets wait for the backup and let me know if any concerns or issues.
- Rancy
Will this affect my actions? - Yes sir reason being the Database are now in Split brain situation so only way out of it is Reseed those Databases to Passive node.
Yeah if Backup is taken today with Replication broken for some DB for whatever reason once i fix and reseed it will purge logs as its criteria of "Copy and replay to passive DB" is met :)
Lets wait for the backup and let me know if any concerns or issues.
- Rancy
Providence_Healthcare: Just want to understand if i wasnt able to answer your queries as Suspend and Resume in your case wouldnt work when your moving logs as the Database and Logs sync breaks between nodes ?
- Rancy
- Rancy
ASKER
We have now had successful backups every night since last week.
Thanks
Thanks
ASKER
- Utilize the Suspend-StorageGroupCopy to stop replication.
- Enable circular logging
- Dismount & remount the storage group
- Start the server service
- Utilize the Resume-StorageGroupCopy to resume replication?