Error codes 0 - 199

Error 2 indicates the key value already exists in index.

You are trying to add a record to a table that has a unique index and matches an existing record.

Error 3 indicates you could not delete target key value since recbyt does not match associated data record position in index.

Error numbe 3 occurs when you are trying to delete a record not in the database.

Error 4 indicates improper use of buffers during ReWriteRecord(). The target key value could not be found in index meaning there was no deletion performed.

When you get error code 4, consider rebuilding the database.

Error -8 is a sysiocod value when FNOP_ERR (error 12) was caused by conflicting open requests (Server).

ctdbAlterTable() can return this if you are doing a Hot Alter Table and the server can't get an exclusive lock on the file.

Error 12 indicates a file could not be opened. Either because the file does not exist, the file name references an incorrect file name, or the file is locked by another process.

For ISAM functions, Check sysiocod for the system-level error and check isam_fil for the specific file number.

Error 14 indicates that the file appears corrupt at open. This occurs if a file is updated and the disk protocol is set at NOTFORCE and CloseIFile() is not executed.

The files have been updated but not properly closed; they were not processed by automatic recovery so they are in an unknown (inconsistent) state.

Error 17 indicates the data file could not be created. Either no space is available on disk or the file name points to an improper name.

FairCom DB was unable to create a new database file. Another file with the same name could already be in the folder.

If you are using the CTDB API, make sure you have created the necessary folders on the server to match the path you are using. The server won't create folders for you.

Error 21 indicates a file does not have the minimum required record length. When a record is deleted, the database engine needs 5 or 9 bytes to track the record offset of deleted records.

If you get error code 21 when creating a data file, increase the record length to 5 or larger for smaller files and 9 or larger for huge files.

If you get error number 21 from an index rebuild, check that IFIL->dreclen is set the same as reclen in the file header. These two must match.

Error 22 indicates that filno out of range: 0 <= filno < fils, where fils is 2nd parameter. This error may occur if FairCom DB has not been initialized.

Either your InitISAM or the Session you opened has too low of a maximum files limit. There are probably many log files in your database because you crashed with a transaction open, and the total files exceeded the limit. Figure out which limit you hit and raise it to a larger number.

Error 26 indicates the file number (datno, keyno, or filno) is not in use.

You have attempted to interact with a file number that does not correspond with an open file. Your file number may be wrong or the file may have been closed by another connection.

Indicates recbyt exceeds logical end of file. If recbyt is correct, then data file header record may be incorrect. If so, rebuild data file.

You can get error number 30 when you are trying to modify a record low-level using an offset that is past the last FairCom DB record in the file even though you are still within the physical file length. You have to add the records before you access them.

Error 34 indicates that the correct predecessor node could not be found.

Error code 34 should not show up in a single user system. It can also indicate that an index insertion was interrupted before completion. In this case, rebuild the index.

Error 35 indicates a seek error, check the sysiocod error.

lseek() failed in function ctio() (ctclib.c). Possible causes are that there is no disk space, there is a corrupted record position in file, or a corrupted file descriptor. It is possible that something is wrong with the local drive's permissions or disk space, or the drive is corrupt.

Error 36 indicates a read error, check the sysiocod error.

The read failed in function ctio() (ctclib.c). This is possibly caused by a corrupted data record position in file. It is possible that something is wrong with the local drive's permissions or disk space, or the drive is corrupt.

Error 39 indicates there are no more records available. The 4-byte data record position (or node position) address space has been exhausted.

To add more records, you can convert it to a large file.

Error 40 indicates the index nose size is too large, and was larger when the index was created. For client/server applications, check your PAGE_SIZE setting in ctsrvr.cfg. For stand-alone applications, check your sect setting (recall that PAGE_SIZE = sect * 128, where sect is the third InitCTree() parameter).

At least one of the indexes has an index node size larger than your FairCom DB connection parameters.

Error number 40 can be avoided by rebuilding the indexes with a smaller node size.

Error 42 indicates that the data record lock could not be obtained, and there is a locked record or table.

Error code 42 could also indicate replication, meaning the record on the target is already locked by another client. Check to see if a "rogue" replication is accidentally running in the background. Troubleshoot replication to fix this issue.

Error 43 indicates a file version error, meaning there is a version compatibility issue. The current configuration parameters are inconsistent with the configuration parameters at the time of file creation.

Error number 43 is also returned by calling functions that require structures with versions, such as GetFileRegions(). These versions must match the versions written into the file in question, otherwise this error is thrown.

Error 46 indicates the file number is already in use. The FILES limit may have been reached; ctsrvr.cfg can be edited to allow a bigger number.

Error number 46 can be thrown from PermIIndex() (PRMIIDX()) after opening a file after the one that is being modified. This function requires a file number sequentially after the host index file number for each item you are adding.

The FILES setting can be increased in the server (or the fils parameter of INITISAM() for standalone). In releases after V10.3, the default for MAX_FILES_PER_USER has been increased from 2,048 to 32,767 to avoid file opens failing due to the lower limit.

Error 48 indicates that the operation is incompatible with the file type. For example, you can use "getRecordsByTable" to return a cursor on a table containing variable-length records. If you fetch a previous record from the cursor, "getRecordsByTable" returns error 48. If you need to fetch records backward, create a cursor based on an index, such as "getRecordsByIndex", "getRecordsByPartialKeyRange", "getRecordsInKeyRange", and "getRecordsStartingAtKey".

Error code 48 can also indicate that the function has been called for the wrong file type (for example, a variable-length function is called for a fixed-length data file).

You can get error number 48 from ctdbLastRecord() if you call it on a file without an index available. ctdbFirstRecord() and ctdbNextRecord() can both operate without an index. ctdbLastRecord() needs an index.

Error 57 indicates the proper lock is not held (ctCHECKLOCK/READ).

In normal ISAM situations, this error means you tried to save a record without having locked it.

Error 64 indicates a checkpoint file is pointing to a region beyond the end of the transaction log.

Error code 64 can also indicate a serious hardware failure caused the final entries of the transaction log to not be written to disk.

When error number 64 occurs you should try restarting the server with the CHECKPOINT_PREVIOUS_YES keyword in place.

Error 64 can also indicate that the log files do not match the database. Check to see if some, but not all, log files were thrown out.

Error 66 indicates the log file entry failed to find a checkpoint.

Prior to V11.6.1, checkpoints that were larger than 16MB were considered bad and this error was thrown. A very large forward roll could create a checkpoint with a very large redo list, causing it to be too large.

Error 67 indicates the file could not be renamed.

Error code 67 can occur when you try to rebuild on a drive other than the system drive. Copy to the system drive and try again. In V9 and earlier, there were issues with rebuilding on a drive other than the startup drive.

Error 69 indicates the node does not belong to index.

Error number 69 could also indicate index corruption caused by things like a server and a single-user library modifying the same tables simultaneously.

Error 70 indicates a transaction is already pending.

Error code 70 can occur when two people are looking at the same info.

A call to PUTHDR() with a mode of ctTIMEIDhdr or ctUNIQIDhdr for a transaction-controlled file that has been updated (LOG_FILE_ID_CHANGE YES) in the current transaction fails with error 70.

Error 71 indicates there is no active transaction.

You are trying to modify a record in a transaction-controlled table without having a transaction open.

Error 86 indicates a deadlock was detected. The FairCom server automatically detects deadlocks, chooses one operation to fail, and returns this error to the connection with the failed operation.

When a connection gets this error, it should retry the operation.

A deadlock may occur when two or more connections attempt to operate on the same resource simultaneously, such as updating the same record.

A deadlock may occur when two or more connections simultaneously attempt to alter or delete different database objects, such as tables and indexes. Because all data objects are stored in the same data dictionary, a deadlock may occur on the dictionary.

Error 94 indicates a transaction could not save or commit a transaction.

Error number 94 can be returned from CommitTransaction() when saving a transaction and from SetSavePoint() when saving a transaction set point. Error 94 may occur when you ignore an error like Keydup and then try to commit the transaction.

If you have a 2-phase commit source/target replicated server, you can get error code 94 when you are trying to connect to a target server as if it were the master. Check to see if you reversed the target and master IP numbers.

Error 96 indicates a start-file/log open error.

This means FairCom DB is having trouble opening a needed log file probably for auto recover. The log could be missing or it could be locked by another process. Two processes trying to open the same single-user database exclusive can result in this error.

If replication is getting this error, it may have gotten so far behind that the max logs was hit and the needed log was deleted. You can re-sync database.

If the Deferred Indexer is logging this error, the logs were likely cleared out but the DFRKSTATEDT.FCS and DFRKSTATEIX.FCS were not deleted and it's still trying to process an old log that is gone.

Error 99 indicates a conflict between the transaction type and mode.

Error code 99 occurs if you try to save a new record, or delete an existing record outside of a transaction.

ICUR_ERR

02000

Error 100 indicates that there is no current ISAM record for the data file isam_fil.

In DirectSQL, error 100 can occur when the current cursor is out of records. It is similar to error 101. In other words, it means data is not found.

Error 101 indicates the ISAM search request for index isam_fil could not be satisfied, and the key does not match any record.

Error 114 indicates the first byte of fixed-length data record found by ISAM routine equals delete flag.

Error 114 can happen if you are trying to delete a record in a fixed-length record that is already deleted, or if a fixed-length record has binary data in the first byte that mimics the deleted mark.

Error 123 indicates a variable-length data record is not preceded by a valid record mark. This means the file is corrupted.

The fixed-length portion length you gave to FairCom DB is incorrect. Check the IFIL Minimum Record Length field, it needs to be the offset of the beginning of the vFields field.

Error number 123 can also indicate corruption in the data file. FairCom was traversing your file and there was other data where a record should have started. This is most often the section of the file that is corrupted.

A rebuild will often fix this, however the intelligence that determines how much data to remove from the data file may cause large numbers of good records to be eliminated. Earlier versions of the rebuild would do this automatically. Newer versions ask.

Always rebuild on a copy and compare the record count before and after.

Error 127 indicates the request could not be sent. Something happened to the server or the network that prevented the client from receiving info from the server. This error is often innocuous. An automatic retransmission most likely solved this.

Review Client/Server ctntio Communication Errors (Formerly VDP Errors) in the FairCom DB documentation for more information.

Error 128 indicates an answer could not be received. Something happened to the server or network that prevented the client from receiving info from the server. This error is often innocuous. An automatic retransmission most likely solves this.

Review Client/Server ctntio Communication Errors (Formerly VDP Errors) in the FairCom DB documentation for more information.

Error 133 indicates that the server could not be identified.

The client was not able to locate the server. You can check to see that the server is working properly and client IP is correct.

Error 147 indicates that the space to be reused is not marked deleted. The record pointed to by available space information is not marked deleted in variable-length data file.

FairCom DB reports this as a diagnostic code. Notify FairCom and ignore.

Error 149 indicates the variable-length passed to AddVRecord() is less than the minimum record length established at file creation.

Error 149 is most often caused by a DODA that is not correct.

Error 160 indicates a key value changed sometime between an index search and the subsequent data record read. This is likely due to multi-user interference, and as a result, the index node for a record did not match the record.

Error code 160 can also indicate that you have an out-of-date index that contains entries for records that no longer exist in the database. A rebuild will clear the bad index entries.

This can happen if you look up a record and another user deletes that record, or if you look up a record and another user edits the record (possibly by making it larger or moving it elsewhere on the disk). In addition, your cached index entry no longer points to the record and will result in this error.

You can catch error 160 and re-read the record to correct this issue, as this will return the corrected offset.

Error 162 indicates the server has gone away.

This could mean that you were connected and the connection was lost. Error 162 can also be returned instead of error 133, for example, when trying to connect to a server that is not there.

Error 179 indicates the server is at full user capacity.

FairCom DB server is reporting you went over the user limit. You can reboot and inform FairCom as this is not supposed to happen.