DSXchange

WOW. Saw ex-CEO of Microsoft say he actually likes Linux on Facebook. Shocked.

I agree with Ray. I never store XML in tables. I would prefer to parse the XML and then write to a table. Then I can format it to whatever is need down the line. In a table you can also query it. It is easy to resend something. It is easy to reconcile. How many rows did I process on Tuesday? Lots of...

Try is as a CHAR or VARCHAR then convert to decimal.

If all the paths are the same to the engine and the projects then this should work. If you changed the paths to the projects by editing UV.ACCOUNT records then that is your problem. The only other issue that could bite you is the mount points are considered B NFS mounts or some other kind of network...

May need to add these environment variables to the dsenv file. You need to stop and restart DataStage engine when you change dsenv.

Otherwise you need to follow Criag's instructions so we can understand more.

Not the same query in the 2 examples. The first example: from table 1 lst inner join table2 ent ON ent.ENT_NO=convert(varchar(9),lst.key) and ent.CUR_FLG = 'Y' inner join table3 ctry on ent.key = ctry.key and ctry.CUR_FLG='Y' inner join table4 on ent.ENT_NO = convert(varchar(9),dps.key) and dps.FLAG...

PRINT statement should be avoided. I have left them in routines by accident. Normally it prints to the log. Create a routine and try it.

If you use generated code then add a quote character to properties then it will get the correct quote character. I am not sure what Oracle's quote character is but it appears not to be a double quote otherwise this would work. Most databases work this same way. Keywords can be column names if quoted...

Agreed.

Not a lot of benefits for separate mount points if on the same SAN if they share the same physical drives. The only benefit is the size of the disk space is reduced. The SAN itself tries to spread the load and most are not optimized for ETL. There is usually lots of buffering in the SAN which helps ...

In one of the releases they converted these to the recommended stages as they imported the jobs. You might try exporting and importing and see if it upgrades your jobs. Save you a lot of time fixing them one at a time. IBM might have something to upgrade them as well. Call support.

If you run a job that runs in 5 minutes with a 1x1 in DEV with 10,000 rows. Then you run same job with 8x8 and it runs in 2 minutes. Does not have much to do when you run 10 million rows through it in PROD. Does not mean it will only be 2.5 times faster. It might be closer to 8 times faster. It also...

EtlStats had a way to get row counts on every job. We used this to see when the last time a job ran to do a similar removal of old jobs. Any jobs not ran in the last 45 days was assumed to be a candidate. That way if a job runs monthly then you are safe. If a job runs annually then you might have is...

I have a script which does this. Here are the dsjob calls in that script: GetLogJobEnd=`$DSPath/dsjob -lognewest $ProjectName $GetLogJobName 2>/dev/null | awk -F= '{print $2}'` GetLogJobBeg=$( $DSPath/dsjob -logsum -type STARTED -max 2 $ProjectName $GetLogJobName 2>/dev/null | head -1 | awk '{print ...

Instance id is a field in the log table. So you can extract it by instance id.