Fix for segfault on prometheus metrics logging

[ilya-maltsev]

Based on #508

[ilya-maltsev]

changes only in sources/cron.c....
sources/frontend.c - realy need to be formatted?

[reshke]

changes only in sources/cron.c.... sources/frontend.c - realy need to be formatted?
Yes, seems that.

If this is true resolves issue, we can actually drop function od_logger_write_plain completely

[ilya-maltsev]

after this changes prometheus metrics in syslog looks like:

Jul 14 06:02:49 test-odyssey odyssey: 1716060 2023-07-14T06:02:49Z info [none-none] [none-none] stats # HELP client_pool_total Total database clients count\n# TYPE client_pool_total gauge\nclient_pool_total{database\="default_db"} 1\n\n# HELP avg_tx_count Average transactions count per second\n# TYPE avg_tx_count gauge\navg_tx_count{user\="default_user",database\="default_db"} 0\n\n# HELP avg_tx_time Average transaction time in usec\n# TYPE avg_tx_time gauge\navg_tx_time{user\="default_user",database\="default_db"} 0\n\n# HELP avg_query_count Average query count per second\n# TYPE avg_query_count gauge\navg_query_count{user\=

but size of prom_metrics more then OD_LOGLINE_MAXLEN

odyssey/sources/logger.h

Line 10 in de02985

#define OD_LOGLINE_MAXLEN 1024

and with od_log all of prometheus metrics cannot be placed in logfile

so I would like to suggest to remove this functionality which was added in earlier implementations of prometheus metrics support - #346