Recent content by George Potemkin

Workaround for OS-COMMAND is found: the extra quotation marks around the whole command: DEFINE VARIABLE vCommand AS CHARACTER NO-UNDO. ASSIGN vCommand = SEARCH("_progres.exe"). OS-COMMAND VALUE('"' + QUOTER(vCommand) + ' -param "a b""'). It's still a mystery.

4 days ago Tom Bascom wrote: https://community.progress.com/s/question/0D5Pb00003uKdhKKAS/pro2-replproc-logging "I wouldn't be using PROWIN.EXE for a background service. That is the job of _PROGRES.EXE" Progress community appears to be divided into parties - the supporters of prowin and the...

Thank you, Patrick! I agree - OS-COMMAND seems to be a dead end. The bug is very tricky. I can't understand its logic. It's the only reason why I'm still trying to find out a solution.

I need to start a Progress session using a OS-COMMAND. There is a space in DLC path , so I need to put the path to the executable file in quotation marks. Also I need to use the quotes for the parameters passed to a Progress executable (e.g. for –pf). But if the command uses more than one pair...

Update to complete my mini investigation of resource queues. Rarely used resource queues: --- DB Buf Backup a.k.a. DBBK --- Db block to be backed up. Not used/not implemented. --- DB Buf Avail a.k.a. BUFF a.k.a. NOBF --- Replacement buffer was needed for the LRU chain but all buffers in the...

The client sessions running r-codes put the share schema locks. RPLS would gain a share schema lock as well. Database restart would close a session that holds the exclusive schema lock but it did not help to restore the default logins. Nevertheless I did check the dates of schema changes in...

Valery Bashkatov adds: Test environment logging-level=3 [2026/03/05@11:56:17.036+0300] P-32070 T-139756064773952 I RPLS : (-----) Session beginning ... [2026/03/05@11:56:18.038+0300] P-32070 T-139756064773952 I RPLS : (10500) The Fathom Replication Server successfully started...

I forgot to mention that "limited logins" allows to start promon. It would be interesting to know if it's possible to identify "limited logins" inside promon (as we can do for a proquiet point).

Progress OpenEdge Release 11.7 build 2182 SP15 on Linux 35 replication server startups per 2025. 7 times it started with limited logins. I tried my best to find out the patterns but I see nothing. DB Size: 7205 GB Size of the AI files locked while RPLS was not running is the hundreds GBs.

Scenario: Replication server is stopped on a source database for a few hours (4-20 hours). When we start a replication server again it may write to a database log: (-----) Enabled limited logins. instead of: (-----) Enabled default logins. It's V11.7. I guess since V12.8 the messages are...

Three lists of wait names from V9.1A executable: REC SCHE TRAN RGET DBRD DBWR DBBK BKSH BKEX BIRD BIWR AIWR IDLE IDLE TXS TXB TXX TXE BUFF REC SCH TRAN DBSI RGET DBRD DBWR DBBK BKSH BKEX BIRD BIWR AIRD AIWR IDLE IDLE TXE TXE REC SCH TRAN DBSI RGET DBRD DBWR DBBK BKSH BKEX BIRD BIWR...

Hi Patrice, I see the number "3" after "TXE" and "OE10" in your data. Some old Progress versions reported the TXE lock waits as _Connect-Wait = "TXE" and _Connect-Wait1 = 1, 2, 3. My "collection" of "Blocked Clients" in promon: 4137 root REMC/ABL TXE 3 20513147...

I would like to programmatically bind _Connect-Wait to _Resrc-Name. Basically _Connect-Wait returns a short name of resource queue a process is waiting for. _Resrc-Name is a full name of the same resource. So the list of the names would be enough. Both match to the wait codes: %GSYSTEM ERROR...

There appear to be two error message stacks: &SCOPED-DEFINE MaxNumber 32767 DEFINE VARIABLE i AS INTEGER NO-UNDO. DEFINE VARIABLE e AS CHARACTER NO-UNDO. /* Put a message in the message stack */ ASSIGN _MSG = {&MaxNumber}. /* without NO-ERROR - in the stack: */ DO ON ERROR UNDO, LEAVE: ASSIGN...

There are two signals that we can control our response to inside a Progress session: SIGINT (2) - Terminal interrupt signal. = Ctrl-C SIGPIPE (13) - Write on a pipe with no one to read it. The SIGPIPE signal causes a session to issue the message 140: ** Pipe to subprocess has been broken...