Does the user have access to the database  RWE_2020_02921?   The exception below seems to indicate he/she does not: " AuthorizationException: User 'zhengzhg' does not have privileges to access: RWE_2020_02921.*"   What is the exact query that is being run against the RWE_2020_02921 database? ... View more

datastream_sender_timeout_ms is an Impala startup flag. See https://impala.apache.org/docs/build/html/topics/impala_config_options.html for information on setting startup flag options.   The "RPC client failed to connect" means that a client failed to connect to an impalad process running on p1i-hdp-srv09.lnt.com. It is hard to say which client failed to make the connection without more information, are there additional client and impalad logs available? What version of Impala is running? Is KRPC enabled on your cluster? ... View more

Unlikely to be related to admission control. The error you included is referred to as a DATASTREAM_SENDER_TIMEOUT error in Impala and is thrown by Impala's RPC layer.   What version of Impala are you running? This could be related to  IMPALA-6818 which reports a similar error. You could try increasing the value of datastream_sender_timeout_ms to a higher value such as 3600000. ... View more

The exchange operators seem to be the bottleneck: 16:EXCHANGE 1 26m14s 26m14s 22.55K 0 0 0 KUDU(KuduPartition(shift_timekey)) 13:EXCHANGE 1 26m13s 26m13s 21.46K -1 0 0 HASH(t.pnl_id,d.oper_code,d.factory) 12:EXCHANGE 1 24m17s 24m17s 108.37M -1 0 0 BROADCAST Each take 20+ minutes. More importantly, the profile says "WARNING: The following tables are missing relevant table and/or column statistics. rptpid.i_f_r_mes_hist_pnl_mod, rptpid.i_f_t_mes_hist_gop_inout_fab". You should compute stats on these tables. Having accurate statistics is important for Impala performance. ... View more

https://impala.apache.org/docs/build/html/topics/impala_admission.html is probably a good place to start. ... View more

https://impala.apache.org/docs/build/html/topics/impala_misc_functions.html#misc_functions__get_json_object has some good documentation on how to use the get_json_object function.   What Impala version are you using? What is the type of the column that contains the JSON data?   I hit a couple issues when parsing the JSON you posted. I believe the JSON standard does not allow for single quotes. Standard online JSON parsers have trouble with the 'u' character as well.   I was able to get the following to work on Impala master:   [localhost:21000] default> select get_json_object("{\"CH4: NO2 (we-aux)\": {\"unit\": \"mV\", \"value\": 4.852294921875}, \"CH6: Ox concentration (we-aux)\": {\"unit\": \"ppb\", \"value\": -84.73094995471016}}", '$.*'); Query: select get_json_object("{\"CH4: NO2 (we-aux)\": {\"unit\": \"mV\", \"value\": 4.852294921875}, \"CH6: Ox concentration (we-aux)\": {\"unit\": \"ppb\", \"value\": -84.73094995471016}}", '$.*') +--------------------------------------------------------------------------------------------------------------------------------------------------------------------------+ | get_json_object('{"ch4: no2 (we-aux)": {"unit": "mv", "value": 4.852294921875}, "ch6: ox concentration (we-aux)": {"unit": "ppb", "value": -84.73094995471016}}', '$.*') | +--------------------------------------------------------------------------------------------------------------------------------------------------------------------------+ | [{"unit":"mV","value":4.852294921875},{"unit":"ppb","value":-84.73094995471016}] | +--------------------------------------------------------------------------------------------------------------------------------------------------------------------------+   ... View more

Offset means the offset into the actual csv file. So in this case, that means the  2432696320th byte of the file foo_042019.csv. There are multiple tools that should allow you to open the file and seek to the desired offset. For example, you could open the file in vim and run :goto 2432696320  which should seek the cursor to the  2432696320th byte of the file, and thus the offending row. ... View more

The error " Error converting column: 35 to TIMESTAMP" means there was an error when converting column 35 to the TIMESTAMP type.   The error "Error parsing row: file: hdfs://blabla/foo_042019.csv, before offset: 2432696320" means there was an error while parsing the row at file offset 2432696320, in the file foo_042019.csv.   So it looks like there are several rows in your dataset where certain fields cannot be converted to TIMESTAMPs. You should be able to open up the file, and seek to the specified offset to find the rows that are corrupted.   I believe, Hive does not throw an exception when given the same dataset, instead it converts the corrupted rows to NULL. The same behavior can be emulated in Impala by setting 'abort_on_error=false'. However, be warned that setting this option can mask data corruption issues. See https://impala.apache.org/docs/build/html/topics/impala_abort_on_error.html for details. ... View more

This looks like an application issue (or possibly an Impala connector issue). The error message states that the Impala query was cancelled by another thread. Likely, another thread in your application got a handle to the query and cancelled it.   The full runtime profile of the query would help as well. ... View more

IMPALA-8557 has been fixed. The fix should be coming in a CDH release soon. ... View more

It depends on how you are updating your partitions. Are you creating completely new partitions, or adding files to existing partitions? ALTER TABLE RECOVER PARTITIONS is specifically used for adding newly created partition directories to a partitioned table - https://impala.apache.org/docs/build/html/topics/impala_alter_table.html ... View more

Queries are in the "waiting to be closed" stage if they are in the EXCEPTION state or if all the rows from the query have been read. In either case, the query needs to be explicitly closed for it to be "completed".   https://community.cloudera.com/t5/Support-Questions/Query-Cancel-and-idle-query-timeout-is-not-working/td-p/58104 might be useful as well. ... View more

The " InactiveTotalTIme" for the "KrpcDataStreamSender" corresponds to the amount of time spent waiting for in-flight RPCs to complete. RPCs are considered complete when they receive a response from the remote server, or if the RPC ends prematurely for other reasons (e.g. cancellation).   Is you attach the full runtime profile of the query, it might be easier to debug. ... View more

The "TransmitData() to X.X.X.X:27000 failed" portion of the error message is thrown by the Impala RPC code. the "Connection timed out (error 110)" is a TCP error. TCP error code 110 corresponds to the error "Connection timed out". So, as the error message states, there was a TCP connection timeout between two Impala processes.   It's hard to debug without more information. What query was being run? Can you post the full log files? What were the two processes that were trying to communicate with each other?   In all likelihood, this looks like a network issue, does this happen consistently? ... View more

I believe OpenLDAP should work. Have you tried using "-- ldap_ca_certificate" instead of "--ca_cert". According to Impala, "--ldap_ca_certificate" is "The full path to the certificate file used to authenticate the LDAP server's certificate for SSL / TLS connections."   Do the logs for the impalad you are trying to connect to contain any relevant debugging information?   The main restriction I am aware of is lack of support for LDAP search / bind operations in Impala - https://issues.apache.org/jira/browse/IMPALA-2563 ... View more

https://www.cloudera.com/documentation/enterprise/5-16-x/topics/impala_ldap.html might be helfpul here.   Have you taken a look at the impalad logs on cdhmaster1? ... View more

Is the stack trace the same every time this crashes? If so, there is likely a bug somewhere in your UDAF, it's hard for me to say exactly where the bug is, but I would recommend trying to reproduce the issue outside Impala (perhaps run the UDAF in a test harness). If you can reproduce the crash outside Impala, standard C++ debugging tools should help you pin down the issue. ... View more

Could you attach the full hs_err.log file? There should be a stack trace in the hs_err.log file of the thread that crashed. The impalad.INFO files would probably help as well. ... View more

The " Connection reset by peer" message should have just been printed by the client (e.g. Hue, impala-shell, etc.). If the impalad itself crashed it should have printed some exit message or other logs that indicated why the impalad crashed (logs should be in impalad.INFO). If the impalad logs don't have anything useful, then it should have written either a minidump or core dump which can be parsed to see what caused it to crash.   How big are your input files? Is it possible that with only one input file, only a single scan range was used to read the table, and that with more input files, multiple scan ranges were spawned? If that is the case, it sounds like there might be a race condition somewhere . ... View more

Agree with the comment from  Tomas79, the filter will convert the LEFT JOIN to an INNER JOIN. I found this post helpful in explaining the behavior: http://wiki.lessthandot.com/index.php/WHERE_conditions_on_a_LEFT_JOIN   You mentioned that things work correctly in Hive, what version of Hive are you using? I checked apache/hive master branch and it follows the same behavior as Impala. ... View more