util: rewritten pilerpurge in python

Change-Id: I2dfa02fa287a02c636f5c91914c3b8e4bef0e764
Signed-off-by: SJ <sj@acts.hu>
This commit is contained in:
SJ 2017-05-14 22:18:05 +02:00
parent c0308ac70b
commit a94019f09f
4 changed files with 189 additions and 10 deletions

View File

@ -33,7 +33,7 @@ MAKE = `which make`
INSTALL = @INSTALL@
all: libpiler.a piler pilerconf pilerget pileraget pilerimport pilerexport pilerpurge reindex test piler-smtp
all: libpiler.a piler pilerconf pilerget pileraget pilerimport pilerexport reindex test piler-smtp
install: install-piler
@ -62,9 +62,6 @@ pilerimport: pilerimport.c libpiler.a
pilerexport: pilerexport.c libpiler.a
$(CC) $(CFLAGS) $(INCDIR) $(DEFS) -o $@ $< -lpiler $(LIBS) $(LIBDIR)
pilerpurge: pilerpurge.c libpiler.a
$(CC) $(CFLAGS) $(INCDIR) $(DEFS) -o $@ $< -lpiler $(LIBS) $(LIBDIR)
pilerconf: pilerconf.c libpiler.a
$(CC) $(CFLAGS) $(INCDIR) $(DEFS) -o $@ $< -lpiler $(LIBS) $(LIBDIR)
@ -91,12 +88,11 @@ install-piler:
$(INSTALL) -m 6755 -o $(RUNNING_USER) -g $(RUNNING_GROUP) pileraget $(DESTDIR)$(bindir)
$(INSTALL) -m 6755 -o $(RUNNING_USER) -g $(RUNNING_GROUP) pilerimport $(DESTDIR)$(bindir)
$(INSTALL) -m 6755 -o $(RUNNING_USER) -g $(RUNNING_GROUP) pilerexport $(DESTDIR)$(bindir)
$(INSTALL) -m 6755 -o $(RUNNING_USER) -g $(RUNNING_GROUP) pilerpurge $(DESTDIR)$(bindir)
$(INSTALL) -m 6755 -o $(RUNNING_USER) -g $(RUNNING_GROUP) reindex $(DESTDIR)$(bindir)
$(INSTALL) -m 6755 -o $(RUNNING_USER) -g $(RUNNING_GROUP) pilertest $(DESTDIR)$(bindir)
clean:
rm -f *.o *.a libpiler.so* piler pilerconf pilerget pileraget pilerimport pilerexport pilerpurge pilertest reindex piler-smtp
rm -f *.o *.a libpiler.so* piler pilerconf pilerget pileraget pilerimport pilerexport pilertest reindex piler-smtp
distclean: clean
rm -f Makefile

View File

@ -42,6 +42,7 @@ install:
$(INSTALL) -m 0755 $(srcdir)/indexer.attachment.sh $(DESTDIR)$(libexecdir)/piler
$(INSTALL) -m 0755 $(srcdir)/import.sh $(DESTDIR)$(libexecdir)/piler
$(INSTALL) -m 0755 $(srcdir)/purge.sh $(DESTDIR)$(libexecdir)/piler
$(INSTALL) -m 0755 $(srcdir)/pilerpurge.py $(DESTDIR)$(libexecdir)/piler
$(INSTALL) -m 0755 $(srcdir)/postinstall.sh $(DESTDIR)$(libexecdir)/piler
$(INSTALL) -m 0755 $(srcdir)/db-mysql.sql $(DESTDIR)$(datarootdir)/piler

184
util/pilerpurge.py Executable file
View File

@ -0,0 +1,184 @@
#!/usr/bin/python
import ConfigParser
import MySQLdb as dbapi
import StringIO
import argparse
import getpass
import os
import sys
import syslog
import time
# Purging of emails works as follows:
#
# 1. Find from metadata table which piler IDs we should remove
#
# 2. Remove the .m files, and set deleted=1 for for those piler IDs
# in metadata table
#
# 3. Find non-referenced attachments, and remove them from filesystem
# then remove them from attachment table
SQL_PURGE_SELECT_QUERY = "SELECT piler_id, size FROM " +\
"metadata WHERE deleted=0 AND retained < UNIX_TIMESTAMP(NOW()) " +\
"AND id NOT IN (SELECT id FROM rcpt WHERE `to` IN " +\
"(SELECT email FROM legal_hold)) AND id NOT IN (SELECT " +\
"id FROM metadata WHERE `from` IN (SELECT email FROM " +\
"legal_hold))"
opts = {}
def read_options(filename="", opts={}):
s = "[piler]\n" + open(filename, 'r').read()
fp = StringIO.StringIO(s)
config = ConfigParser.RawConfigParser()
config.readfp(fp)
opts['username'] = config.get('piler', 'mysqluser')
opts['password'] = config.get('piler', 'mysqlpwd')
opts['database'] = config.get('piler', 'mysqldb')
opts['storedir'] = config.get('piler', 'queuedir')
opts['server_id'] = "%02x" % config.getint('piler', 'server_id')
def purge_m_files(ids=[], opts={}):
if len(ids) > 0:
remove_m_files(ids, opts)
if opts['dry_run'] is False:
# Set deleted=1 for remove metadata entries
cursor = opts['db'].cursor()
format = ", ".join(['%s'] * len(ids))
cursor.execute("UPDATE metadata SET deleted=1 WHERE piler_id IN " +
"(%s)" % (format), ids)
opts['db'].commit()
def purge_attachments(ids=[], opts={}):
format = ", ".join(['%s'] * len(ids))
cursor = opts['db'].cursor()
# Select non referenced attachments
cursor.execute("SELECT i, piler_id, attachment_id FROM v_attachment " +
"WHERE refcount=0 AND piler_id IN (%s)" % (format), ids)
while True:
rows = cursor.fetchall()
if rows == ():
break
remove_ids = []
# Delete attachments from filesystem
for (id, piler_id, attachment_id) in rows:
remove_ids.append(id)
if opts['dry_run'] is False:
unlink(get_attachment_file_path(piler_id, attachment_id, opts))
else:
print get_attachment_file_path(piler_id, attachment_id, opts)
# Delete these IDs from attachment table
if opts['dry_run'] is False:
format = ", ".join(['%d'] * len(remove_ids))
cursor.execute("DELETE FROM attachment WHERE piler_id IN (%s)" %
(format), remove_ids)
opts['db'].commit()
else:
print remove_ids
def remove_m_files(ids=[], opts={}):
for i in range(0, len(ids)):
if opts['dry_run'] is False:
unlink(get_m_file_path(ids[i], opts), opts)
else:
print get_m_file_path(ids[i], opts)
def unlink(filename="", opts={}):
try:
st = os.stat(filename)
opts['purged_stored_size'] = opts['purged_stored_size'] + st.st_size
opts['count'] = opts['count'] + 1
os.unlink(filename)
except:
pass
def get_m_file_path(id='', opts={}):
return "/".join([opts['storedir'], opts['server_id'], id[8:11], id[32:34],
id[34:36], id + ".m"])
def get_attachment_file_path(piler_id='', attachment_id=0, opts={}):
return "/".join([opts['storedir'], opts['server_id'], piler_id[8:11],
piler_id[32:34], piler_id[34:36], piler_id + ".a" +
str(attachment_id)])
def main():
parser = argparse.ArgumentParser()
parser.add_argument("-c", "--config", type=str, help="piler.conf path",
default="/usr/local/etc/piler/piler.conf")
parser.add_argument("-b", "--batch-size", type=int, help="batch size " +
"to delete", default=1000)
parser.add_argument("-d", "--dry-run", help="dry run", action='store_true')
parser.add_argument("-v", "--verbose", help="verbose mode",
action='store_true')
args = parser.parse_args()
if getpass.getuser() not in ['root', 'piler']:
print "Please run me as user 'piler'"
sys.exit(1)
opts['dry_run'] = args.dry_run
opts['db'] = None
opts['count'] = 0
opts['size'] = 0
opts['purged_size'] = 0
opts['purged_stored_size'] = 0
syslog.openlog(logoption=syslog.LOG_PID, facility=syslog.LOG_MAIL)
read_options(args.config, opts)
try:
opts['db'] = dbapi.connect("localhost", opts['username'],
opts['password'], opts['database'])
cursor = opts['db'].cursor()
cursor.execute(SQL_PURGE_SELECT_QUERY)
while True:
rows = cursor.fetchmany(args.batch_size)
if rows == ():
break
piler_id = [x[0] for x in rows]
size = [x[1] for x in rows]
opts['purged_size'] = opts['purged_size'] + sum(size)
purge_m_files(piler_id, opts)
purge_attachments(piler_id, opts)
except dbapi.DatabaseError, e:
print "Error %s" % e
if opts['db']:
opts['db'].close()
syslog.syslog("purged " + str(opts['count']) + " files, " +
str(opts['purged_size']) + "/" +
str(opts['purged_stored_size']) + " bytes")
if __name__ == "__main__":
main()

View File

@ -1,6 +1,6 @@
#!/bin/bash
export PATH=/bin:/sbin:/usr/bin:/usr/sbin:/usr/local/bin:/usr/local/sbin
export PATH=/bin:/sbin:/usr/bin:/usr/sbin:/usr/libexec/piler:/usr/local/libexec/piler
PRIORITY=mail.error
TMPFILE=/var/run/piler/purge.tmp
PURGE_BEACON=/var/piler/stat/purge
@ -17,6 +17,4 @@ trap finish EXIT
touch $PURGE_BEACON
pilerpurge
pilerpurge.py